**Diabetic Foot Ulceration and Amputation**

## Stephanie Burns1 and Yih-Kuen Jan2

*1Veterans Affairs Medical Center, Department of Physical Therapy, 2University of Oklahoma Health Sciences Center, Department of Rehabilitation Sciences, Oklahoma City, Oklahoma, USA* 

#### **1. Introduction**

The number of people with diabetes mellitus (DM) has been conservatively estimated to approximately double by 2030 to a worldwide prevalence of 4.4% at which time 366 million people will have diabetes (Wild et al., 2004). As the number of people with DM rises, so too will the burden of diabetic foot disease, particularly since the factors contributing to ulcer formation such as peripheral neuropathy and vascular disease are already present in 10% of people at the time of diagnosis (Boulton et al., 2005). The risk of an individual with DM developing a foot ulcer some time in his or her lifetime could be as high as 15% and foot ulcers are found in 12% to 25% of diabetics (Singh et al., 2005; Brem et al., 2006). Results from population and community based studies in the UK have shown a 1.3-4.8% prevalence rate of foot ulcers in persons with type 2 DM (Boulton et al., 2005). The annual incidence of foot ulceration is more than 2% among all persons with diabetes and 5% to 7.6% among diabetics with peripheral neuropathy (Abbott et al., 2002; Boulton et al., 2004).

The prevalence of diabetes-related complications such as peripheral neuropathy and foot disease will continue to increase in countries such as the United States not only as the prevalence of the disease increases but as longevity of the population with DM improves. Among people with DM, lower extremity disease is the most common source of complications and hospitalization (Boyko et al.). Ghanassia et al (2008) reported a diabetic foot ulcer recurrence rate of 60.9% and an amputation rate of 43.8% in a study of 89 hospitalized subjects (Ghanassia et al., 2008). Almost 50% of nontraumatic lower extremity amputations worldwide occur in people with DM (Global Lower Extremity Amputation Study, 2000). Amputations from complications related to DM place an individual at risk for additional amputation and have a 5 year mortality rate of 39% to 68% (Morris et al., 1998). People with diabetic foot ulcers have a lower health-related quality of life than the general population and diabetics without foot ulcers as well (Ribu et al., 2007).

#### **2. Pathophysiology of diabetic foot ulceration**

The pathogenesis of diabetic foot ulceration is multifactorial and the result of a complex interplay of a number of elements including peripheral neuropathy, structural deformities, elevated plantar pressures, limited joint mobility, vascular disease, and various extrinsic sources of trauma such as ill fitting shoe wear or foreign objects in shoes. The peripheral

Diabetic Foot Ulceration and Amputation 3

Nearly 50% of all people with DM have diabetic polyneuropathy making it one of the most common long-term complications of the disease with chronic, symmetrical, sensorimotor polyneuropathy being the most typical type (Tesfaye et al., 2010). Persons with DM and signs of peripheral neuropathy have been shown to be 4 times as likely to have plantar ulcerations as those without neuropathy (Frykberg et al., 1998). Presence of peripheral neuropathy induces a number of pathologic changes in the diabetic foot that then interact to increase susceptibility to ulceration. Sensory neuropathy can affect perception of pain, pressure, touch, temperature, and proprioception. Loss of protective sensation prevents detection of levels of injurious trauma to tissue and stimuli that would ordinarily trigger a protective response such as ill fitting footwear or a foreign object in a shoe go unperceived, often until extensive destruction has occurred. Loss of sensation has been shown to be associated with diabetic foot ulceration in a number of studies (Boyko et al., 1999; Reiber et al., 1999). Results of a prospective multicenter study point to sensory neuropathy as the most frequent component in the causal sequence to diabetic foot ulceration (Reiber et al., 1999). Proprioceptive loss leads to instability and changes in gait that can increase the

As polyneuropathy progresses, motor fibers are affected resulting in weakness and atrophy of the distal leg and intrinsic foot muscles (Andreassen et al., 2006). Motor neuropathy can lead to foot deformities such as claw or hammertoes, prominent metatarsal heads, or hallux valgus. Prevalence of clawing or hammering toes in persons with DM has been reported to be 32 to 46% (Holewski et al., 1989; Smith et al., 1997). Hammer toe is an important predictor of plantar pressure (Mueller et al., 2003) and claw/hammer toe deformity is associated with elevated plantar pressures at the MTHs (Bus et al., 2005). Intrinsic foot muscle weakness has long been thought to be a proximate cause of deformity in the diabetic foot (Reiber et al., 1999). The intrinsic muscles of the foot ordinarily function to balance the pull of the extrinsic flexors and extensors at the interphalangeal joints by flexing the MTP joints while extending the interphalangeal joints. Weakness of the intrinsic muscles leads to loss of this stabilizing function and ultimately hyperextension of the MTP joints and clawing of the toes. Fat pads underlying the metatarsal heads, embedded in the flexor tendons and originating from the plantar ligaments attached to the proximal phalanges, tend to migrate distally when the toes claw resulting in removal of the soft tissue cushion beneath the metatarsal heads. The prominent metatarsal heads are now exposed to abnormally high plantar pressures during walking as plantar tissue thickness has been shown to be related to peak plantar pressures (Abouaesha et al., 2001). Findings of two recent studies have raised questions about the causal relationship between muscle atrophy and deformity noting that intrinsic muscle atrophy was present before clinical peripheral neuropathy could be detected and finding no significant difference in degree of intrinsic foot muscle atrophy between matched subjects

with and without claw toe deformity (Greenman et al., 2005; Bus et al., 2009).

Concomitant damage to the sympathetic fibers in peripheral neuropathy results in sudomotor dysfunction that can trigger a cascade of untoward effects in the foot beginning with atrophy of the sweat glands and progressing through anhidrosis, drying of the skin, fissuring and callus formation (Vinik et al., 2003). Excessive drying has been associated with foot ulceration (Tentolouris et al., 2009). Foot temperature increases in parallel with a reduction in sweating and this may predispose to infection (Sun et al., 2008; O'Loughlin et

**2.2 Diabetic polyneuropathy and ulceration** 

potential for traumatic injury.

neuropathy that occurs in DM is truly a "poly"neuropathy in that sensory, motor and autonomic fibers and function are all adversely affected. It is the sequelae of these neural dysfunctions in conjunction with extrinsic factors that produce the physiologic and structural changes that lead to ulceration. The most common causal pathway to diabetic foot ulceration involves the confluence of loss of sensation resulting in failure to detect repetitive pressure or trauma and abnormal foot structure or deformity producing sites of abnormally high pressure, usually over areas of bony prominence (Mueller et al., 1990; Brem et al., 2006; Chao and Cheing, 2009; O'Loughlin et al., 2010). Diabetic peripheral polyneuropathy is the central component as it can induce changes in foot structure and produce dryness of the skin which can lead to callus formation (van Schie, 2006; O'Loughlin et al., 2010). Callosities form on areas of elevated pressure on the plantar aspect of the foot in response to pressure amplified by restricted joint motion of the ankle and foot which is applied to dry, poorly lubricated skin resulting from autonomic dysfunction (Young et al., 1992). Loss of protective sensation permits continuation of repetitive pressure that goes undetected causing calluses to thicken into sources of tissue trauma then hemorrhage and ulcerate underneath (Murray et al., 1996). Veves et al. (1992) first demonstrated the relationship between high plantar pressures and diabetic foot ulceration in a prospective study in 1992. The relative risk of developing an ulcer in an area of high plantar pressure is 4.7 and that risk more than doubles to 11.0 at the site of a callus (Murray et al., 1996).

#### **2.1 Types of diabetic foot ulcers**

Diabetic foot ulcers are classified as one of 3 types based on their primary etiologies and clinical characteristics: neuropathic, neuroischemic, and ischemic. This classification is a reflection of the physiological systems adversely impacted by the chronic hyperglycemia of the disease. Hyperglycemia induces alterations in multiple metabolic pathways resulting in structural and functional changes in the microvasculature of local tissue and the peripheral nerves in cases of peripheral neuropathy (Chao and Cheing, 2009). Neuropathic ulcers appear in the absence of protective sensation as a result of peripheral sensory neuropathy but without evidence of macrovascular disease. The presence of co-morbidity, deep foot infection, and plantar or metatarsal head ulcer location have been shown to be related to minor and major amputation risk in diabetic patients without ischemia (Gershater et al., 2009). They are typically found on the plantar surfaces of the feet and make up about 40% of all diabetic foot ulcers.

Diabetic foot ulcers are considered vascular or ischemic in origin when they occur in the absence of palpable pedal pulses (posterior tibial and dorsalis pedis arteries) in conjunction with ankle brachial indices (ABIs) of less than 0.9. Infection is coincident with ischemia in 50% of patients with this type of diabetic foot ulcer (Dinh et al.; Prompers et al., 2007). This type of ulcer comprises about 10% of all diabetic foot ulcerations. As their name implies, neuroischemic ulcers share features common to both ischemic and neuropathic ulcers in that they occur in the absence of protective sensation and palpable pedal pulses. They make up the final 40% of diabetic foot ulcers. Probability of major amputation in diabetic patients with ischemic/neuroischemic ulcers has been related to the extent of peripheral vascular disease, presence of co-morbidity, multiple ulcerations and tissue loss (Gershater et al., 2009). Peripheral vascular disease is the most important factor related to outcome in these types of diabetic foot ulcers (Boulton et al., 2005; Gershater et al., 2009).

neuropathy that occurs in DM is truly a "poly"neuropathy in that sensory, motor and autonomic fibers and function are all adversely affected. It is the sequelae of these neural dysfunctions in conjunction with extrinsic factors that produce the physiologic and structural changes that lead to ulceration. The most common causal pathway to diabetic foot ulceration involves the confluence of loss of sensation resulting in failure to detect repetitive pressure or trauma and abnormal foot structure or deformity producing sites of abnormally high pressure, usually over areas of bony prominence (Mueller et al., 1990; Brem et al., 2006; Chao and Cheing, 2009; O'Loughlin et al., 2010). Diabetic peripheral polyneuropathy is the central component as it can induce changes in foot structure and produce dryness of the skin which can lead to callus formation (van Schie, 2006; O'Loughlin et al., 2010). Callosities form on areas of elevated pressure on the plantar aspect of the foot in response to pressure amplified by restricted joint motion of the ankle and foot which is applied to dry, poorly lubricated skin resulting from autonomic dysfunction (Young et al., 1992). Loss of protective sensation permits continuation of repetitive pressure that goes undetected causing calluses to thicken into sources of tissue trauma then hemorrhage and ulcerate underneath (Murray et al., 1996). Veves et al. (1992) first demonstrated the relationship between high plantar pressures and diabetic foot ulceration in a prospective study in 1992. The relative risk of developing an ulcer in an area of high plantar pressure is 4.7 and that risk more than

Diabetic foot ulcers are classified as one of 3 types based on their primary etiologies and clinical characteristics: neuropathic, neuroischemic, and ischemic. This classification is a reflection of the physiological systems adversely impacted by the chronic hyperglycemia of the disease. Hyperglycemia induces alterations in multiple metabolic pathways resulting in structural and functional changes in the microvasculature of local tissue and the peripheral nerves in cases of peripheral neuropathy (Chao and Cheing, 2009). Neuropathic ulcers appear in the absence of protective sensation as a result of peripheral sensory neuropathy but without evidence of macrovascular disease. The presence of co-morbidity, deep foot infection, and plantar or metatarsal head ulcer location have been shown to be related to minor and major amputation risk in diabetic patients without ischemia (Gershater et al., 2009). They are typically found on the plantar surfaces of the feet and make up about 40% of

Diabetic foot ulcers are considered vascular or ischemic in origin when they occur in the absence of palpable pedal pulses (posterior tibial and dorsalis pedis arteries) in conjunction with ankle brachial indices (ABIs) of less than 0.9. Infection is coincident with ischemia in 50% of patients with this type of diabetic foot ulcer (Dinh et al.; Prompers et al., 2007). This type of ulcer comprises about 10% of all diabetic foot ulcerations. As their name implies, neuroischemic ulcers share features common to both ischemic and neuropathic ulcers in that they occur in the absence of protective sensation and palpable pedal pulses. They make up the final 40% of diabetic foot ulcers. Probability of major amputation in diabetic patients with ischemic/neuroischemic ulcers has been related to the extent of peripheral vascular disease, presence of co-morbidity, multiple ulcerations and tissue loss (Gershater et al., 2009). Peripheral vascular disease is the most important factor related to outcome in these

types of diabetic foot ulcers (Boulton et al., 2005; Gershater et al., 2009).

doubles to 11.0 at the site of a callus (Murray et al., 1996).

**2.1 Types of diabetic foot ulcers** 

all diabetic foot ulcers.

#### **2.2 Diabetic polyneuropathy and ulceration**

Nearly 50% of all people with DM have diabetic polyneuropathy making it one of the most common long-term complications of the disease with chronic, symmetrical, sensorimotor polyneuropathy being the most typical type (Tesfaye et al., 2010). Persons with DM and signs of peripheral neuropathy have been shown to be 4 times as likely to have plantar ulcerations as those without neuropathy (Frykberg et al., 1998). Presence of peripheral neuropathy induces a number of pathologic changes in the diabetic foot that then interact to increase susceptibility to ulceration. Sensory neuropathy can affect perception of pain, pressure, touch, temperature, and proprioception. Loss of protective sensation prevents detection of levels of injurious trauma to tissue and stimuli that would ordinarily trigger a protective response such as ill fitting footwear or a foreign object in a shoe go unperceived, often until extensive destruction has occurred. Loss of sensation has been shown to be associated with diabetic foot ulceration in a number of studies (Boyko et al., 1999; Reiber et al., 1999). Results of a prospective multicenter study point to sensory neuropathy as the most frequent component in the causal sequence to diabetic foot ulceration (Reiber et al., 1999). Proprioceptive loss leads to instability and changes in gait that can increase the potential for traumatic injury.

As polyneuropathy progresses, motor fibers are affected resulting in weakness and atrophy of the distal leg and intrinsic foot muscles (Andreassen et al., 2006). Motor neuropathy can lead to foot deformities such as claw or hammertoes, prominent metatarsal heads, or hallux valgus. Prevalence of clawing or hammering toes in persons with DM has been reported to be 32 to 46% (Holewski et al., 1989; Smith et al., 1997). Hammer toe is an important predictor of plantar pressure (Mueller et al., 2003) and claw/hammer toe deformity is associated with elevated plantar pressures at the MTHs (Bus et al., 2005). Intrinsic foot muscle weakness has long been thought to be a proximate cause of deformity in the diabetic foot (Reiber et al., 1999). The intrinsic muscles of the foot ordinarily function to balance the pull of the extrinsic flexors and extensors at the interphalangeal joints by flexing the MTP joints while extending the interphalangeal joints. Weakness of the intrinsic muscles leads to loss of this stabilizing function and ultimately hyperextension of the MTP joints and clawing of the toes. Fat pads underlying the metatarsal heads, embedded in the flexor tendons and originating from the plantar ligaments attached to the proximal phalanges, tend to migrate distally when the toes claw resulting in removal of the soft tissue cushion beneath the metatarsal heads. The prominent metatarsal heads are now exposed to abnormally high plantar pressures during walking as plantar tissue thickness has been shown to be related to peak plantar pressures (Abouaesha et al., 2001). Findings of two recent studies have raised questions about the causal relationship between muscle atrophy and deformity noting that intrinsic muscle atrophy was present before clinical peripheral neuropathy could be detected and finding no significant difference in degree of intrinsic foot muscle atrophy between matched subjects with and without claw toe deformity (Greenman et al., 2005; Bus et al., 2009).

Concomitant damage to the sympathetic fibers in peripheral neuropathy results in sudomotor dysfunction that can trigger a cascade of untoward effects in the foot beginning with atrophy of the sweat glands and progressing through anhidrosis, drying of the skin, fissuring and callus formation (Vinik et al., 2003). Excessive drying has been associated with foot ulceration (Tentolouris et al., 2009). Foot temperature increases in parallel with a reduction in sweating and this may predispose to infection (Sun et al., 2008; O'Loughlin et

Diabetic Foot Ulceration and Amputation 5

(PAD) is estimated to occur twice as frequently among persons with DM as those without (Dinh et al.). Lower extremity arterial insufficiency in persons with DM can have both macro- and microvascular components. Probability of healing in diabetic foot ulcers has been shown to be strongly related to severity of peripheral vascular disease (Apelqvist et al., 2011). The reported prevalence of PAD in patients with diabetic foot ulcers ranges from 10% to 60% (Armstrong and Lavery, 1998; Oyibo et al., 2001; Moulik et al., 2003). A multi-center trial in Europe reported an overall PAD prevalence of 49% but this varied from 22 to 73% among various centers (Prompers et al., 2007). Peripheral arterial disease typically affects infrapopliteal vessels specifically the profunda femoris in people with DM (Dinh et al.).

Tissue viability ultimately depends on adequate local blood supply to cells via the microcirculation. Alterations in microcirculation have been implicated in formation of diabetic foot ulcers for some time (Dinh and Veves). Dysfunction in the microcirculation of the diabetic foot is not occlusive in nature but secondary to structural and functional changes (Dinh et al.; Chao and Cheing, 2009). The chronic hyperglycemia brought on by DM leads to intracellular accumulation of glucose inducing alterations in multiple metabolic pathways in vascular and neural tissue. Hyperglycemia is a causative factor in impaired vascular permeability and tone as well as auto regulation of blood flow (Chao and Cheing, 2009). Impaired vasodilatory response to plantar pressure causing tissue ischemia is the common final pathway, according to various theories, of the development of diabetic foot ulcers (Boulton et al., 2000). Diabetic patients (with or without peripheral neuropathy) suffer from various forms of microvascular dysfunction, including abnormal vasomotion (Benbow et al., 1995; Stansberry et al., 1996; Bernardi et al., 1997), impaired vasodilatory response to local heating (Malik et al., 1993; Stansberry et al., 1999), decreased blood flow under or after pressure loading (Fromy et al., 2002; Koitka et al., 2004), endothelial nitric oxide dysfunction (Veves et al., 1998), and attenuated response to sympathetic maneuvers (Aso et al., 1997).

Thickening of basement membranes and reduction in capillary size are structural changes that are more prominent in the lower extremities (Dinh et al.). Functionally, vasoreactivity is impaired via reduction in both endothelium-dependent and non-endothelium dependent vasodilation. Both endothelium- and non-endothelium-dependent vasodilation are impaired in the presence of peripheral neuropathy while PAD primarily affects non-endotheliumdependent vasodilation (Dinh et al.; Veves and King, 2001). Occlusive vascular lesions would be more amenable to surgical intervention while the functional ischemia resulting from dysfunctional vasoreactivity would be less responsive to bypass procedures (Veves et al., 1998). Therefore correction of macrocirculatory issues will not necessarily result in

healing of a diabetic foot ulcer or prevention of one in the future (Arora et al., 2002).

sensory neurons for substance P and CGRP reduced (Levy et al.; Caselli et al., 2003).

Microcirculation in persons with DM can also be adversely affected by the neuropathic impairment of the nerve-axon reflex. Stimulation of the C-nociceptive nerve fibers ordinarily leads to release of local vasodilators such as substance P, bradykinin and calcitonin generelated peptide (CGRP). These neuropeptides act to produce vasodilation via direct action on vascular smooth muscle or indirectly on mast cells through histamine release. This axon mediated response normally accounts for roughly 1/3 of the endothelium-dependent vasodilation in the foot and forearm (Hamdy et al., 2001). This neurogenic vasodilatory response is impaired in the presence of diabetic peripheral neuropathy and the number of

al., 2010). Tentolouris et al. (2009) found sudomotor dysfunction was associated with an almost 15 times greater risk of foot ulceration and similarly Sun et al. (2008) reported the risk of plantar ulceration occurrence was 13.4 times greater in a patient group with the most sudomotor dysfunction over a 4 year follow-up period.

#### **2.3 Biomechanical factors and ulceration**

Limited motion at the ankle or limited joint mobility has been associated with increased peak forefoot pressures and risk of ulceration and re-ulceration (Delbridge et al., 1988). The exact pathogenesis of limited joint mobility in DM is unclear but it is thought to be due to progressive stiffening of the collagen-containing tissues ultimately resulting in thickening of the skin with loss of joint motion (Zimny et al., 2004). Giacomozzi and colleagues demonstrated reduced ankle mobility in patients with DM with and without peripheral neuropathy suggesting another mechanism is responsible for alterations in foot-ankle biomechanics (Giacomozzi et al.). Abnormal thicknesses of plantar fascia and Achilles tendon have been measured (D'Ambrogi et al., 2005; Salsich et al., 2005).

Alterations in biomechanical properties of the diabetic foot have been proven to cause increased plantar foot pressure, which may lead to the development of diabetic foot ulcers (Mueller et al., 2003). Diabetes is associated with the formation of glucose-mediated intermolecular cross-links (i.e. advanced glycation end-products, AGE). Accumulations of AGEs increase stiffness of the cartilages, muscles, tendons, ligaments, and skin (Brownlee et al., 1988). A stiffer plantar soft tissue reduces the shock-absorbing mechanism of the anklefoot complex and may make the diabetic foot more vulnerable to repetitive stress during walking (Landsman et al., 1995).

The hallux has been identified as the most common site of diabetic foot ulceration, accounting for 20% to 30% of diabetic foot ulcers in a study of 360 patients and comprising 22% of the ulcers seen in another research group's clinic (Armstrong et al., 1998; Nube et al., 2006). Several risk factors have been associated with ulceration of the hallux. Decreased dorsiflexion at the first metatarsophalangeal joint, neuropathy, increased length of the hallux, increased interphalangeal angle, increased body weight, decreased soft tissue thickness and pes planus are all associated with increased pressure at the hallux (Mueller et al., 2003).

Another common deformity seen in diabetic feet is Charcot's neuroarthropathy. Charcot's foot is characterized by neuropathic fractures of the midfoot region resulting in collapse of the arch of the foot. Involvement of the tarsal joints can cause the plantar surface to become convex resulting in the classic "rocker-bottom" foot. This deformity leads to areas of elevated pressure on skin that is not adapted to tolerate pressure and ultimately leads to ulceration (Mueller et al., 1990). Abnormal perfusion of the bones of the midfoot precipitated by autonomic neuropathy may be an etiologic component (O'Loughlin et al., 2010). Both Charcot deformity and hammer toes have been shown to be independent risk factors for diabetic foot ulcers (Boyko et al., 1999).

#### **2.4 Microvascular factors and ulceration**

Adequate vascular supply is essential for healing and ischemia often plays a role in ulceration of the diabetic foot. Wound healing requires an adequate supply of oxygen and nutrients be provided to cells involved in the repair process. Peripheral arterial disease

al., 2010). Tentolouris et al. (2009) found sudomotor dysfunction was associated with an almost 15 times greater risk of foot ulceration and similarly Sun et al. (2008) reported the risk of plantar ulceration occurrence was 13.4 times greater in a patient group with the most

Limited motion at the ankle or limited joint mobility has been associated with increased peak forefoot pressures and risk of ulceration and re-ulceration (Delbridge et al., 1988). The exact pathogenesis of limited joint mobility in DM is unclear but it is thought to be due to progressive stiffening of the collagen-containing tissues ultimately resulting in thickening of the skin with loss of joint motion (Zimny et al., 2004). Giacomozzi and colleagues demonstrated reduced ankle mobility in patients with DM with and without peripheral neuropathy suggesting another mechanism is responsible for alterations in foot-ankle biomechanics (Giacomozzi et al.). Abnormal thicknesses of plantar fascia and Achilles

Alterations in biomechanical properties of the diabetic foot have been proven to cause increased plantar foot pressure, which may lead to the development of diabetic foot ulcers (Mueller et al., 2003). Diabetes is associated with the formation of glucose-mediated intermolecular cross-links (i.e. advanced glycation end-products, AGE). Accumulations of AGEs increase stiffness of the cartilages, muscles, tendons, ligaments, and skin (Brownlee et al., 1988). A stiffer plantar soft tissue reduces the shock-absorbing mechanism of the anklefoot complex and may make the diabetic foot more vulnerable to repetitive stress during

The hallux has been identified as the most common site of diabetic foot ulceration, accounting for 20% to 30% of diabetic foot ulcers in a study of 360 patients and comprising 22% of the ulcers seen in another research group's clinic (Armstrong et al., 1998; Nube et al., 2006). Several risk factors have been associated with ulceration of the hallux. Decreased dorsiflexion at the first metatarsophalangeal joint, neuropathy, increased length of the hallux, increased interphalangeal angle, increased body weight, decreased soft tissue thickness and pes planus

Another common deformity seen in diabetic feet is Charcot's neuroarthropathy. Charcot's foot is characterized by neuropathic fractures of the midfoot region resulting in collapse of the arch of the foot. Involvement of the tarsal joints can cause the plantar surface to become convex resulting in the classic "rocker-bottom" foot. This deformity leads to areas of elevated pressure on skin that is not adapted to tolerate pressure and ultimately leads to ulceration (Mueller et al., 1990). Abnormal perfusion of the bones of the midfoot precipitated by autonomic neuropathy may be an etiologic component (O'Loughlin et al., 2010). Both Charcot deformity and hammer toes have been shown to be independent risk

Adequate vascular supply is essential for healing and ischemia often plays a role in ulceration of the diabetic foot. Wound healing requires an adequate supply of oxygen and nutrients be provided to cells involved in the repair process. Peripheral arterial disease

tendon have been measured (D'Ambrogi et al., 2005; Salsich et al., 2005).

are all associated with increased pressure at the hallux (Mueller et al., 2003).

factors for diabetic foot ulcers (Boyko et al., 1999).

**2.4 Microvascular factors and ulceration** 

sudomotor dysfunction over a 4 year follow-up period.

**2.3 Biomechanical factors and ulceration** 

walking (Landsman et al., 1995).

(PAD) is estimated to occur twice as frequently among persons with DM as those without (Dinh et al.). Lower extremity arterial insufficiency in persons with DM can have both macro- and microvascular components. Probability of healing in diabetic foot ulcers has been shown to be strongly related to severity of peripheral vascular disease (Apelqvist et al., 2011). The reported prevalence of PAD in patients with diabetic foot ulcers ranges from 10% to 60% (Armstrong and Lavery, 1998; Oyibo et al., 2001; Moulik et al., 2003). A multi-center trial in Europe reported an overall PAD prevalence of 49% but this varied from 22 to 73% among various centers (Prompers et al., 2007). Peripheral arterial disease typically affects infrapopliteal vessels specifically the profunda femoris in people with DM (Dinh et al.).

Tissue viability ultimately depends on adequate local blood supply to cells via the microcirculation. Alterations in microcirculation have been implicated in formation of diabetic foot ulcers for some time (Dinh and Veves). Dysfunction in the microcirculation of the diabetic foot is not occlusive in nature but secondary to structural and functional changes (Dinh et al.; Chao and Cheing, 2009). The chronic hyperglycemia brought on by DM leads to intracellular accumulation of glucose inducing alterations in multiple metabolic pathways in vascular and neural tissue. Hyperglycemia is a causative factor in impaired vascular permeability and tone as well as auto regulation of blood flow (Chao and Cheing, 2009). Impaired vasodilatory response to plantar pressure causing tissue ischemia is the common final pathway, according to various theories, of the development of diabetic foot ulcers (Boulton et al., 2000). Diabetic patients (with or without peripheral neuropathy) suffer from various forms of microvascular dysfunction, including abnormal vasomotion (Benbow et al., 1995; Stansberry et al., 1996; Bernardi et al., 1997), impaired vasodilatory response to local heating (Malik et al., 1993; Stansberry et al., 1999), decreased blood flow under or after pressure loading (Fromy et al., 2002; Koitka et al., 2004), endothelial nitric oxide dysfunction (Veves et al., 1998), and attenuated response to sympathetic maneuvers (Aso et al., 1997).

Thickening of basement membranes and reduction in capillary size are structural changes that are more prominent in the lower extremities (Dinh et al.). Functionally, vasoreactivity is impaired via reduction in both endothelium-dependent and non-endothelium dependent vasodilation. Both endothelium- and non-endothelium-dependent vasodilation are impaired in the presence of peripheral neuropathy while PAD primarily affects non-endotheliumdependent vasodilation (Dinh et al.; Veves and King, 2001). Occlusive vascular lesions would be more amenable to surgical intervention while the functional ischemia resulting from dysfunctional vasoreactivity would be less responsive to bypass procedures (Veves et al., 1998). Therefore correction of macrocirculatory issues will not necessarily result in healing of a diabetic foot ulcer or prevention of one in the future (Arora et al., 2002).

Microcirculation in persons with DM can also be adversely affected by the neuropathic impairment of the nerve-axon reflex. Stimulation of the C-nociceptive nerve fibers ordinarily leads to release of local vasodilators such as substance P, bradykinin and calcitonin generelated peptide (CGRP). These neuropeptides act to produce vasodilation via direct action on vascular smooth muscle or indirectly on mast cells through histamine release. This axon mediated response normally accounts for roughly 1/3 of the endothelium-dependent vasodilation in the foot and forearm (Hamdy et al., 2001). This neurogenic vasodilatory response is impaired in the presence of diabetic peripheral neuropathy and the number of sensory neurons for substance P and CGRP reduced (Levy et al.; Caselli et al., 2003).

Diabetic Foot Ulceration and Amputation 7

Chronic hyperglycemia and insulin use, which could be considered a marker for glycemic control, have been shown to be independent risk factors for LEA in persons with DM (Adler et al., 1999; Davis et al., 2006; Adler et al., 2010). Elevated HbA1c is associated with risk of LEA such that for every 1% increase in HbA1c there is an associated 26% to 36% increased risk of LEA (Adler et al., 2010). Positive associations have been observed between glycemia and micro- and macrovascular complications and clinical trials have demonstrated the value of improved glycemic control on microvascular complications (DCCT, 1993; UKPDS, 1998). Data on macrovascular complications and glycemic control is less clear with limited clinical trial data to unequivocally demonstrate that intensive glycemic control reduces risk of LEA

Increased risk of LEA associated with hyperglycemia is thought to be mediated by PAD and peripheral sensory neuropathy. Various biochemical changes resulting from hyperglycemia including glycation, protein kinase C activation, sorbitol and hexosamine pathway activation result in arterial disease, sensory neuropathy, autonomic dysfunction and ultimately deregulation of blood flow (Adler et al., 2010). History of diabetic foot ulcers and previous amputation are both independent predictors of LEAs (Adler et al., 1999; Resnick et al., 2004; Davis et al., 2006). Presence of a diabetic foot ulcer is the single biggest risk factor for nontraumatic amputation in persons with DM and increases the risk of amputation 6 fold (Brem et al., 2006; Davis et al., 2006). A diabetic foot ulcer precedes 85% of major LEAs in individuals with DM (Larsson et al., 1997). The presence of a diabetic foot ulcer alone in a person with DM increases the risk of LEA 7 times relative to patients with Charcot arthropathy alone and diabetic foot ulcers together with Charcot arthropathy increases the

(Zoungas et al., 2008; Patel et al., 2009; Adler et al., 2010).

risk of LEA 12 times versus Charcot arthropathy alone (Sohn et al., 2010).

**2.7 Morbidity and mortality following diabetes-related lower extremity amputation** 

The causal factors leading to the initial amputation remain in place following LEA and continue to place these individuals at elevated risk for re-ulceration. Re-ulceration risk is higher in those with a previous amputation due to increased pressure on a smaller residual weight bearing area, abnormal pressure distribution on the remaining plantar surface and alterations in bony architecture. Thirty-four percent of amputees re-ulcerate in the first year and 70% after 5 years (Apelqvist et al., 1993). Further amputation is twice as likely in persons with DM than in those without with 22% undergoing another amputation a median of 7 months following initial amputation (Schofield et al., 2006). Re-amputation at a higher level on the residual limb is a function of disease progression, failure to heal, and risk factors that develop as a result of the initial amputation such as alteration in the pressure distribution on the residual weight bearing surface. Age and heel lesions have also been shown to be risk factors for re-amputation (Skoutas et al., 2009). Risk of re-amputation is highest within the first 6 months of initial amputation (Izumi et al., 2006; Skoutas et al., 2009). A re-amputation rate of 21.5% within 18 months was reported by Skoutas et al (2009) and 1 year and 3 year rates of 26.7% and 48.3% by Izumi (2006). Forty percent of subjects with DM in a study by Tentolouris et al. had an ipsilateral or contralateral amputation within an average of about 16 months of the first DM-related LEA (Tentolouris et al., 2004). Mortality risk following LEA is higher for individuals with DM than those without DM. People with DM had a 55% increased risk of death after amputation compared to those without DM (Schofield et al., 2006). One of the first prospective studies on long-term

#### **2.5 Diabetic foot ulcers and lower extremity amputation**

DM increases the risk for lower extremity amputation (LEA) from 2% to 16% depending on study design and the population studied (Adler et al., 1999; Lavery et al., 2003; Resnick et al., 2004; Frykberg et al., 2006). Rates of LEA among persons with DM can be as much as 15 to 40 times higher than their non-DM counterparts (Lavery et al., 1996; Resnick et al., 1999). Incidence rates of all LEAs are 4-7 times higher in men and women with DM than in people without DM (Frykberg et al., 2006). A Dutch study found the incidence rate of initial unilateral LEA was 8 times higher in persons with DM than in persons without DM (Johannesson et al., 2009). Lavery et al. found men with DM were 2.35 times more likely to have an LEA than women with DM (Lavery et al., 1999). In a Native American population with DM, risk of LEA was twice as high for men as women (Resnick et al., 2004). Amputation risk varies among ethnic groups being 1.72 to 2.17 times higher in African Americans than non-Hispanic whites and Hispanics (Lavery et al., 1996) and Native Americans, Hispanic Americans and African Americans having a 1.5 to 2.4 fold increased risk of DM-related LEAs than their age-matched Caucasian counterparts (Lavery et al., 1999; Resnick et al., 2004).

The majority of LEAs due to DM were toe amputations followed by BKAs then AKAs and foot amputations with rates of 2.6, 1.6, and 0.8 per 1000 in 2002 (Centers for Disease Control and Prevention, 2005). Several studies in the US and western Europe in recent years have reported decreasing incidence of LEAs in DM populations particularly in response to implementation of improved diabetes foot care (Krishnan et al., 2008; Schofield et al., 2009). In the 5 year longitudinal study by Canavan et al. (2008), the incidence rate of LEA in persons with DM dropped from 310.5 per 100,000 persons to 75.9 per 100,000. A similar dramatic 62% reduction in incidence of major LEAs and a more modest 40.3% decline in total LEAs over 11 years were reported (Krishnan et al., 2008). However, a large retrospective study utilizing a nationwide sample in England found no significant decrease in incidence of DM-related LEAs from 2004 to 2008 (Vamos et al., 2010). The explanation for the differences in findings may lie in the differences in study design as retrospective studies have been reported to underestimate incidence by 4.2% to 90.6% and misclassify 4.5% to 17.4% of amputations (Rayman et al., 2004).

#### **2.6 Risk factors for diabetes-related amputation**

Generally speaking, the same factors involved in ulceration of the diabetic foot can have at least contributory roles in LEAs. PAD, infection, chronic hyperglycemia, and history of previous diabetic foot ulcers or amputation are significant risk factors for amputation. Ischemia is a contributory if not the major factor determining the need for a LEA (Schofield et al., 2006). PAD is an independent risk factor for LEA in people with DM (Adler et al., 1999; Moulik et al., 2003; Davis et al., 2006). Adequate blood supply is necessary for healing and resolution of infection as impaired blood interferes with tissue oxygenation and antibiotic delivery to affected regions. PAD is present in 8% of adults with DM at the time of diagnosis and there is a 3.5 fold risk among men with DM and a 8.6 fold risk among women of developing PAD (Melton et al., 1980; Kannel, 1985). In a study by Moulik et al. (2003), 59% of patients who had LEAs over a 5 year follow-up period had PAD and 5 year amputation rates were higher and times to amputation were shorter in this group. While infection may not be an independent risk factor for LEA is often related to inadequate blood flow and interferes with healing (Reiber et al., 1999).

DM increases the risk for lower extremity amputation (LEA) from 2% to 16% depending on study design and the population studied (Adler et al., 1999; Lavery et al., 2003; Resnick et al., 2004; Frykberg et al., 2006). Rates of LEA among persons with DM can be as much as 15 to 40 times higher than their non-DM counterparts (Lavery et al., 1996; Resnick et al., 1999). Incidence rates of all LEAs are 4-7 times higher in men and women with DM than in people without DM (Frykberg et al., 2006). A Dutch study found the incidence rate of initial unilateral LEA was 8 times higher in persons with DM than in persons without DM (Johannesson et al., 2009). Lavery et al. found men with DM were 2.35 times more likely to have an LEA than women with DM (Lavery et al., 1999). In a Native American population with DM, risk of LEA was twice as high for men as women (Resnick et al., 2004). Amputation risk varies among ethnic groups being 1.72 to 2.17 times higher in African Americans than non-Hispanic whites and Hispanics (Lavery et al., 1996) and Native Americans, Hispanic Americans and African Americans having a 1.5 to 2.4 fold increased risk of DM-related LEAs than their age-matched Caucasian counterparts (Lavery et al., 1999;

The majority of LEAs due to DM were toe amputations followed by BKAs then AKAs and foot amputations with rates of 2.6, 1.6, and 0.8 per 1000 in 2002 (Centers for Disease Control and Prevention, 2005). Several studies in the US and western Europe in recent years have reported decreasing incidence of LEAs in DM populations particularly in response to implementation of improved diabetes foot care (Krishnan et al., 2008; Schofield et al., 2009). In the 5 year longitudinal study by Canavan et al. (2008), the incidence rate of LEA in persons with DM dropped from 310.5 per 100,000 persons to 75.9 per 100,000. A similar dramatic 62% reduction in incidence of major LEAs and a more modest 40.3% decline in total LEAs over 11 years were reported (Krishnan et al., 2008). However, a large retrospective study utilizing a nationwide sample in England found no significant decrease in incidence of DM-related LEAs from 2004 to 2008 (Vamos et al., 2010). The explanation for the differences in findings may lie in the differences in study design as retrospective studies have been reported to underestimate incidence by 4.2% to 90.6% and misclassify 4.5% to

Generally speaking, the same factors involved in ulceration of the diabetic foot can have at least contributory roles in LEAs. PAD, infection, chronic hyperglycemia, and history of previous diabetic foot ulcers or amputation are significant risk factors for amputation. Ischemia is a contributory if not the major factor determining the need for a LEA (Schofield et al., 2006). PAD is an independent risk factor for LEA in people with DM (Adler et al., 1999; Moulik et al., 2003; Davis et al., 2006). Adequate blood supply is necessary for healing and resolution of infection as impaired blood interferes with tissue oxygenation and antibiotic delivery to affected regions. PAD is present in 8% of adults with DM at the time of diagnosis and there is a 3.5 fold risk among men with DM and a 8.6 fold risk among women of developing PAD (Melton et al., 1980; Kannel, 1985). In a study by Moulik et al. (2003), 59% of patients who had LEAs over a 5 year follow-up period had PAD and 5 year amputation rates were higher and times to amputation were shorter in this group. While infection may not be an independent risk factor for LEA is often related to inadequate blood

**2.5 Diabetic foot ulcers and lower extremity amputation** 

Resnick et al., 2004).

17.4% of amputations (Rayman et al., 2004).

**2.6 Risk factors for diabetes-related amputation** 

flow and interferes with healing (Reiber et al., 1999).

Chronic hyperglycemia and insulin use, which could be considered a marker for glycemic control, have been shown to be independent risk factors for LEA in persons with DM (Adler et al., 1999; Davis et al., 2006; Adler et al., 2010). Elevated HbA1c is associated with risk of LEA such that for every 1% increase in HbA1c there is an associated 26% to 36% increased risk of LEA (Adler et al., 2010). Positive associations have been observed between glycemia and micro- and macrovascular complications and clinical trials have demonstrated the value of improved glycemic control on microvascular complications (DCCT, 1993; UKPDS, 1998). Data on macrovascular complications and glycemic control is less clear with limited clinical trial data to unequivocally demonstrate that intensive glycemic control reduces risk of LEA (Zoungas et al., 2008; Patel et al., 2009; Adler et al., 2010).

Increased risk of LEA associated with hyperglycemia is thought to be mediated by PAD and peripheral sensory neuropathy. Various biochemical changes resulting from hyperglycemia including glycation, protein kinase C activation, sorbitol and hexosamine pathway activation result in arterial disease, sensory neuropathy, autonomic dysfunction and ultimately deregulation of blood flow (Adler et al., 2010). History of diabetic foot ulcers and previous amputation are both independent predictors of LEAs (Adler et al., 1999; Resnick et al., 2004; Davis et al., 2006). Presence of a diabetic foot ulcer is the single biggest risk factor for nontraumatic amputation in persons with DM and increases the risk of amputation 6 fold (Brem et al., 2006; Davis et al., 2006). A diabetic foot ulcer precedes 85% of major LEAs in individuals with DM (Larsson et al., 1997). The presence of a diabetic foot ulcer alone in a person with DM increases the risk of LEA 7 times relative to patients with Charcot arthropathy alone and diabetic foot ulcers together with Charcot arthropathy increases the risk of LEA 12 times versus Charcot arthropathy alone (Sohn et al., 2010).

#### **2.7 Morbidity and mortality following diabetes-related lower extremity amputation**

The causal factors leading to the initial amputation remain in place following LEA and continue to place these individuals at elevated risk for re-ulceration. Re-ulceration risk is higher in those with a previous amputation due to increased pressure on a smaller residual weight bearing area, abnormal pressure distribution on the remaining plantar surface and alterations in bony architecture. Thirty-four percent of amputees re-ulcerate in the first year and 70% after 5 years (Apelqvist et al., 1993). Further amputation is twice as likely in persons with DM than in those without with 22% undergoing another amputation a median of 7 months following initial amputation (Schofield et al., 2006). Re-amputation at a higher level on the residual limb is a function of disease progression, failure to heal, and risk factors that develop as a result of the initial amputation such as alteration in the pressure distribution on the residual weight bearing surface. Age and heel lesions have also been shown to be risk factors for re-amputation (Skoutas et al., 2009). Risk of re-amputation is highest within the first 6 months of initial amputation (Izumi et al., 2006; Skoutas et al., 2009). A re-amputation rate of 21.5% within 18 months was reported by Skoutas et al (2009) and 1 year and 3 year rates of 26.7% and 48.3% by Izumi (2006). Forty percent of subjects with DM in a study by Tentolouris et al. had an ipsilateral or contralateral amputation within an average of about 16 months of the first DM-related LEA (Tentolouris et al., 2004).

Mortality risk following LEA is higher for individuals with DM than those without DM. People with DM had a 55% increased risk of death after amputation compared to those without DM (Schofield et al., 2006). One of the first prospective studies on long-term

Diabetic Foot Ulceration and Amputation 9

(TCC) remains the gold standard for off-loading although removable walkers have also been shown to provide a similar degree of pressure relief. Peak pressure reduction in the forefoot is reported to be up to 87% with TCC but only 44% to 64% with cast shoes and forefoot offloading shoes (Cavanagh and Bus, 2011). Rocker bottom outsoles, custom insoles, metatarsal pads and arch supports may reduce forefoot peak pressure 16% to 52% compared

Effectiveness of an off-loading device must be gauged by both its ability to relieve pressure and patients' adherence to the treatment. TCCs are considered to be effective in part because they essentially coerce patient adherence to treatment. Some of the unloading is achieved by restricting ankle motion and redistributing load to the device itself which may explain why devices that extend only to the ankle are less effective in off-loading the foot than those that reach above the ankle (Cavanagh and Bus, 2011). The majority of evidence for off-loading comes from studies examining uncomplicated neuropathic plantar ulcers. TCC has been shown to be more effective in time to healing than removable devices in some randomized clinical trials while a recent RCT showed similar healing rates between a TCC and an ankle high removable walker (Faglia et al., 2010). Off-loading has been used to treat neuroischemic or infected wounds but success rates are much lower than for purely neuropathic ulcers (Nabuurs-Franssen et al., 2005). TCCs are not in wider use because of potential adverse reactions which include diminished activity level, problems sleeping or driving a car and

Cavanagh and Bus (2011) summarized the recommendations of the International Working Group on the Diabetic Foot for use of off-loading in management of non-complicated foot ulcers in their review: 1) pressure relief should be part of every treatment plan; 2) TCC and non-removable walkers are preferred but clinicians should be aware of potential adverse effects; 3) forefoot off-loading shoes or cast shoes may be used when the above devices are contraindicated or not tolerated; and 4) conventional or standard footwear should not be

Peripheral vascular disease is common in persons with DM and is characterized by impairment at both macro- and microvascular levels. Re-establishing arterial supply is the key to healing ischemic and neuroischemic ulcers. Treatment of peripheral arterial disease involves management of risk factors, medical therapy, and endovascular or open surgery. Smoking cessation, weight loss, and adherence to a low fat diet are all areas in which eliciting patient cooperation is critical for successful management. Antiplatelet therapy, anticoagulation, and LDL lowering drugs may also play a role in treatment. However, many diabetic patients will need re-vascularization to achieve healing. Macrovascular disease is morphologically the same in diabetics and non-diabetics differing only in location with the anterior and posterior tibial and peroneal arteries of the calf being most affected in persons with DM. Surgical options are dependent on whether the vascular disease is supra-inguinal (aorto-iliac) or infra-inguinal (femoro-popliteal-crural) or both ((Ruef et al., 2004). Angioplasty, endoarterectomy, grafting, and by-pass are some available surgical interventions. Vascular surgery may be able to aid in revascularization of an area via restoring flow through larger vessels but will not completely restore the microvascular flow disrupted by structural changes in the basement membranes or functional impairment in

to controls (Cavanagh and Bus, 2011).

iatrogenic ulcers from poorly applied casts.

used as other devices are more effective.

microcirculation caused by the disease.

**3.3 Revascularization**

prognosis following LEA amputation reported 1, 3, and 5 year mortality rates of 15%, 38%, and 68%, respectively for both minor and major amputations combined (Larsson et al., 1997). Almost 10 years later, researchers were still reporting people with DM who underwent LEA had a 55% greater risk of dying than those without DM (Schofield et al., 2006).
