Diabetes mellitus: an important risk factor for peripheral vascular disease

KEYWORDS:

• Diabetes
• peripheral arterial disease
• endovascular revascularization

Peripheral arterial disease (PAD) of the lower extremities affects approximately 8.5 million people in the U.S. and more than 200 million people worldwide [1], while its prevalence is expected to increase [2]. PAD is an independent risk factor for cardiovascular morbidity, including coronary artery disease (CAD) and cerebrovascular accidents (CVA) [3]. Consistent with this, patients with PAD are at almost six-fold higher risk for acute myocardial infarction (MI), ischemic stroke and/or death compared to general population [4,5]. Therefore, studies investigating the prevalence of PAD among high risk groups and identifying risk factors for PAD are crucial in order to optimize the treatment strategies for those patients and improve PAD prognosis. It should be also taken into account that as the U.S. population is highly diverse, specific data on PAD course among several ethnic groups are of significant importance [6]. The American Diabetes Association (ADA) provided evidence that African Americans and Hispanics have the highest prevalence of concomitant PAD and DM [4]. Additionally, a large retrospective analysis from the National Health and Nutrition Examination Survey (NHANES) investigating the trends of traditional PAD risk factors among several ethic groups demonstrated that the prevalence of hypertension (HTN), dyslipidemia, diabetes mellitus (DM) and smoking were associated with the patient ethnicity, with non-Hispanic African Americans being more likely to have traditional PAD risk factors [7].

Traditional risk factors for PAD include but are not limited to age, active smoking, HTN, dyslipidemia, chronic kidney disease, and DM [8,9]. However, as smoking rates in the Western countries are decreasing, DM is expected to become one of the most significant contributors to the development and progression of PAD. Previous studies have demonstrated that up to 30% of patients with PAD have DM, however taking into account the asymptomatic nature of less severe vascular disease and the impaired pain perception among patients with DM, the actual prevalence of DM among patients with PAD may be higher [10–12]. DM is a significant risk factor for atherosclerotic disease [13], especially at below the knee artery segments [14,15] and can occur due to impaired insulin secretion or tissue tolerance to circulated insulin [16]. The pathophysiologic mechanisms of vascular disease in the presence of DM include: inflammation, endothelial cell dysfunction, smooth muscle cell migration, altered platelet function and hyper-coagulability, which eventually lead to foam cell formation, indicating that DM may share similar pathogenic pathways to atherosclerosis and PAD development [4,11,17–21]. Recent evidence regarding the role of DM among patients with PAD is presented in Table 1.

Table 1. Recent evidence regarding peripheral artery disease course among diabetics.

Correlation between diabetes and peripheral artery disease
PAD is more likely to progress among diabetic patients
Diabetes is a major risk factor for PAD development
20%-30% of patients with PAD have diabetes
PAD presents in up to 50% of diabetic patients with diabetic foot ulcer
Glucose intolerance has been correlated with >20% prevalence of abnormal ABI
Patients with coexisting PAD and diabetes are at higher risk for cardiovascular and limb ischemic events, even when being on optimal medical therapy
Diabetes is associated with more diffuse calcified lesions, located mainly at the infrapopliteal segment
Coexistence of diabetes and PAD is associated with higher all-cause mortality, cardiovascular mortality, lower extremity amputations, and limb revascularization rates
Patients with DM and PAD are less likely to be on optimal medical therapy
The prevalence of coexistent diabetes and PAD is higher among patients with CLI (>50%)
Either surgical or endovascular interventions are warranted for the treatment of symptomatic PAD among diabetics
TBI or skin perfusion pressure have been considered the most efficient and most cost-effective approaches for the identification of PAD among diabetic patients with diabetic foot ulcers.
However, the role of DM in PAD is not completely understood and the role of new therapeutic approaches, including optimized medical therapy and new endovascular or surgical modalities is still unclear
Correlation between diabetes and outcomes of endovascular therapy for peripheral artery disease
Diabetes has been correlated with higher 30-day readmission rates after revascularization for PAD
Diabetes may be associated with higher restenosis rates
Diabetes may be associated with higher incidence of cardiovascular outcomes
Diabetes coexistent with CLI may be associated with MALE
As diabetes is associated mainly with below the knee arterial lesions, outcomes of EVT are worse as distal run off diminish
The periprocedural metabolic milieu may play an important role in late restenosis rates
Synchronous PAD in patients with diabetic ulcers might be correlated with lack of wound and higher amputation rates healing
Diabetes and PAD-related amputations could be prevented as a foot ulcer commonly precedes, allowing timely intervention and limb salvage
Optimized anti-thrombotic medical therapy is needed among diabetic patients undergoing EVT for PAD

*DM: diabetes mellitus, PAD: peripheral arterial disease, CLI: critical limb ischemia, TBI: toe-branchial index, MALE: major adverse limb events, EVT: endovascular therapy

Previous studies have provided evidence that DM increases the risk for lower-extremity amputation, especially among patients with critical limb ischemia (CLI) [22,23], leading to higher cardiovascular morbidity/mortality rates compared to non-diabetics [13,16]. Patients with DM who undergo amputation due to PAD have a 50-74% 5-year all-cause mortality, which is mainly associated with cardiac and cerebrovascular events [24]. Furthermore, a recent post-hoc analysis from the ‘Exenatide Study of Cardiovascular Event Lowering’ (EXSCEL; ClinicalTrials.gov Identifier: NCT01144338) study, including 2,800 diabetic patients with/without PAD, evaluated the effects of once per week exenatide (glucagon-like peptide 1 agonist) vs placebo [25]. The EXSCEL study demonstrated that patients with DM and PAD had higher rates of all-cause death, lower-extremity amputation and target limb revascularization compared to patients without PAD, over a median follow up of 3.2 years [25]. Moreover, a large retrospective study including almost 62,300 patients from the Healthcare Cost and Utilization Project Nationwide Inpatient Sample, showed that patients with coexistent PAD and DM had higher prevalence of lower-extremity amputation (47%) compared to patients with either DM alone (26%) or PAD alone (26%), indicating that DM and PAD have an additive effect on overall risk [26].

Thus, inadequate DM care has been associated not only with a higher risk for adverse outcomes (e.g. stroke, death, MI etc.) but also with higher PAD prevalence among diabetics [27]. In up to 85% of DM related amputations an ischemic foot ulcer precedes [28,29]. As such appropriate management with wound care and early vascular intervention at the time of first symptoms or ulceration might eventually prevent/delay the amputation [28,30–32]. Thus, the goal of therapy in addition to avoid complications of DM, is to lower morbidity and mortality rates by preventing/delaying vascular adverse events [32,33]. In general aggressive management of all modifiable risk factors (e.g. smoking cessation, statin therapy, anti-hypertensive therapy, anti-diabetic therapy, healthy diet, supervised/unsupervised exercise etc.) is recommended and when medical and/or exercise therapy fail, surgical or endovascular revascularization is recommended for the treatment of symptomatic PAD among diabetics [10]. A summary of currently available guidelines regarding PAD management is presented in Table 2 [9,34–36].

Table 2. Summary of Current Guidelines for lower extremity artery disease management.

Patients status	SVS 2015 [36]	AHA/ACC 2016 [35]	ESC 2017 [9]	GVG 2019 (CLI) [34]
Asymptomatic PAD	smoking cessation; education about signs and symptoms of PAD progression; noninvasive treatment	smoking cessation; anti-PLT monotherapy when ABI values are abnormal; statin therapy; anti-hypertensive therapy if indicated; diabetes management; annual influenza vaccination; self-foot examination and healthy foot behaviors; considering foot examination by clinician	Anti-PLT should be considered (individualized); statin therapy; smoking cessation; healthy diet and physical activity; strict diabetes management; anti-hypertensive therapy if indicated; ACEI/ARBs should be first-line therapy if PAD and HTN coexist	-
Symptomatic PAD	smoking cessation; diabetes management (HbA1 c < 7%); statin therapy; β-blockers if indicated; anti-PLT monotherapy; against using warfarin solely to reduce cardiovascular events and vascular occlusion; against B12 and folic acid supplements; 3-month trial of cilostazol/pentoxifylline; ACEI ramipril; exercise therapy; EVT or surgical treatment in patients with life-limiting claudication when medical/exercise therapy failed	smoking cessation; diabetes management; monotherapy with aspirin or clopidogrel; statin therapy; anti-hypertensive therapy if indicated; considering ACEI/ARBs; against using anti-coagulation solely to reduce cardiovascular events and vascular occlusion; considering cilostazol (pentoxifylline has no benefit); against B12 supplements; annual influenza vaccination; exercise therapy; self-foot examination and healthy foot behaviors; considering foot examination by clinician; EVT or surgical treatment in patients with life-limiting claudication when medical/exercise therapy failed; revascularization should not be performed solely to prevent progression to CLI	Monotherapy with aspirin or clopidogrel; clopidogrel should be preferred; statin therapy; smoking cessation; healthy diet and physical activity; strict diabetes management; anti-hypertensive therapy if indicated; ACEI/ARBs should be first-line therapy if PAD and HTN coexist; supervised exercise therapy; EVT or surgical treatment in patients with life-limiting claudication when medical/exercise therapy failed; in infrapopliteal CLI cases revascularization is indicated for limb salvage	smoking cessation; monotherapy with aspirin or clopidogrel; considering aspirin and rivaroxaban; manage all modifiable risk factors; against vitamin K antagonists; moderate/high intensity statin therapy; anti-hypertensive therapy if indicated; metformin should be the primary anti-diabetic agent used; stage CLI; revascularization should be offered in all patients with advanced limb-threating conditions based on staging; amputation to unsalvageable limb

*EVT: endovascular therapy, PAD: peripheral artery disease, anti-PLT: antiplatelet therapy, DAPT: dual anti-PLT therapy, CLI: critical limb ischemia, SVS: Society of Vascular Surgery, AHA/ACC: American College of Cardiology/American Heart Association Guideline, ESC: European Society of Cardiology, GVG: Global Vascular Guidelines, MACE: major adverse cardiologic event; ACEI: angiotensin converting enzyme inhibitors; ARB: angiotensin II receptor blocker

Traditionally, DM has been associated with worse outcomes after endovascular interventions, which was mainly attributed to the higher prevalence of infrapopliteal disease among diabetics and consequent inadequate distal runoff [13]. Infrapopliteal lesions are often technically challenging to treat due to longer lesion length, smaller vessel diameter and severe calcification [37]. As such, recoil, dissections, and restenosis remain a limitation of below the knee endovascular interventions [38–41]. Favaretto et al. conducted a prospective study of patients undergoing either iliac or femoropopliteal endovascular revascularization and found that DM and CLI were associated with increased rates of restenosis during a follow up duration of six months [42]. Furthermore, experience from the coronary arteries is also indicative of higher restenosis rates among diabetics. A multivariate analysis from the ‘Rapamycin-Eluting Stent Evaluated At Rotterdam Cardiology Hospital’ (RESEARCH) registry, including almost 100 patients undergoing percutaneous revascularization with sirolimus eluting stents for ST-elevation acute MI, showed that DM was an independent predictor of restenosis [43]. Hereby a previous meta-analysis of six studies, compromising 6,236 patients (1,166 diabetics vs 5,070 non-diabetics) undergoing coronary artery stenting, showed that restenosis occurred more frequently among the DM group [44]. Chronic hyperglycemia promotes the expression of fibroblast growth factor (FGF) and transforming growth factor-α (TGF-α), thereby promoting hyperplasia of arterial wall smooth muscle cells and extracellular matrix production [45]. Additionally, increased oxidative stress and endothelial dysfunction could also explain the high restenosis rates in diabetics undergoing percutaneous transluminal angioplasty [45]. Therefore, as DM contributes to restenosis after percutaneous transluminal interventions [46,47], management of PAD with surgical bypass could be considered more durable [10,48].

However, recent studies have suggested that outcomes after revascularization (either surgical bypass or endovascular therapy) for the treatment of PAD among diabetics with adequate distal runoff vs non-diabetics, were comparable. Hicks et al. studied almost 2,600 patients from the Vascular Quality Initiative (VQI) registry, who underwent infrageniculate bypass or endovascular intervention for CLI, attributed to lesions located at or below the knee [49]. This study showed that infrapopliteal disease causing CLI was successfully treated with surgery or percutaneous transluminal angioplasty (PTA) in both diabetics and non-diabetics, with similar 1-year primary patency (81% vs 79%), major amputation (14% vs 11%) and mortality rates (6% vs 7%) [49]. Additionally, a large multicenter study including 2,566 patients from the OLIVE registry (500 cases underwent surgical revascularization; diabetics: 355 vs non-diabetics: 145; 2,066 cases underwent percutaneous revascularization; diabetics: 1,463 vs non-diabetics: 603) investigated potential predictors of 3-year limb adverse events after endovascular therapy for CLI [50]. The study failed to show any correlations of diabetic status and 3-year major amputation or death [50]. Therefore, the authors suggested that in cases of advanced PAD aggressive revascularization for limb salvage should be attempted independently of DM status [50].

Recently, Lee et al. investigated the 5-year clinical outcomes of patients who underwent PTA for symptomatic PAD among diabetics vs non-diabetics [51]. This study included 560 diabetic and 205 non-diabetic patients with PAD and showed that both groups had similar 5-year cardiac and vascular outcomes [51]. However, patients with DM experienced higher rates of periprocedural major hematoma formation [51]. Furthermore, a recent systematic review and meta-analysis, including overall 14 studies (526,008 patients), demonstrated that after revascularization for PAD diabetics vs non-diabetics were more likely to be readmitted, indicating a potentially higher incidence of periprocedural complications in this group of patients (including bleeding complications) [52]. Therefore, optimal anti-thrombotic regimens (anti-platelets, anti-coagulants) and better vascular access techniques are needed to reduce the risk of periprocedural complications (e.g. bleeding, local hematoma etc.), while glycemic control might also play a significant role in the management of diabetic patients with PAD.

Insulin remains a valuable treatment option for adequate blood glucose control in DM, however its long-term safety is still debatable [53]. Results from previous reports investigating inhibitors of the sodium-glucose cotransporters type-2 (SGLT-2) have shown both adequate glycemic control and cardiovascular mortality benefit [54–56]. However, based on the results from the ‘CANagliflozin cardioVascular Assessment Study’ (CANVAS; ClinicalTrials.gov Identifier: NCT01032629) and ‘A Study of the Effects of Canagliflozin on Renal Endpoints in Adult Participants With Type 2 Diabetes Mellitus’ (CANVAS-R; ClinicalTrials.gov Identifier: NCT01989754) trials concerns were raised due to higher rates of lower extremity amputations in patients with PAD and DM treated with a SGLT-2 inhibitor (2-fold higher incidence of lower limb amputations among patients treated with canagliflozin vs placebo) [57–59]. Due to these findings, the FDA and the European Medicines Agency (EMA) have issued warning notices about higher amputation risk with SGLT-2 inhibitors [60]. However, further clinical trials investigating the higher amputation signal with canagliflozin among diabetics failed to show any difference in limb-loss between canagliflozin vs other non-SGLT-2 agents given for the treatment of DM [61].

Interestingly, a sub-analysis retrieving data from the ‘Empagliflozin Cardiovascular Outcome Event Trial in Type 2 Diabetes Mellitus Patients’ study (EMPA-REG OUTCOME; ClinicalTrials.gov Identifier: NCT01131676), demonstrated superior limb survival rates among patients treated with empagliflozin vs placebo [62]. Additionally, dapagliflozin was found to improve vascular endothelial function and flow-mediated dilation when added to metformin compared to metformin alone [63]. As such, it has been hypothesized that differences in baseline characteristics of the patients enrolled in the several studies investigating the efficacy and safety of SGLT-2 inhibitors might have contributed to this heterogeneity in the reported outcomes of SGLT-2 studies [64]. Thereby, it remains unclear whether the higher amputation signal is a class or individual drug effect and as such further research with head to head comparisons is warranted in order to identify optimal medical therapy for patients with concomitant PAD and DM. Although intensive glycemic control is crucial for the prevention of DM related complications, it has not reduced the rates of cardiovascular adverse events overall in diabetic patients, while controversy still remains regarding the contribution of glycemic control to limb outcomes in diabetic patients undergoing endovascular procedures [65,66].

Singh et al. investigated the short- and long-term outcomes of 309 infrapopliteal lesions (149 patients with DM) treated with balloon angioplasty [27]. The authors suggested that high fasting glucose blood levels at the time of the index procedure were associated with lower primary patency rates and more major adverse limb events (major amputation, new bypass graft placement, surgical graft revision, thrombectomy, thrombolysis) during a mean follow up of 15 months [27]. Similarly, Takahara et al. studying 278 patients with CLI, who underwent PTA, demonstrated that poor glycemic control was correlated with higher amputation rates [22]. In addition, there has been evidence that higher blood glucose levels are associated with worse outcomes after vascular surgery as well [67]. A recent, large comparative analysis among diabetics vs non-diabetics from the ‘Examining Use of Ticagrelor in Peripheral Artery Disease’ (EUCLID; ClinicalTrials.gov Identifier: NCT01732822) trial, which enrolled 13,885 patients with symptomatic PAD, demonstrated that every 1% increase in HbA1 c was associated with a 14.2% increased relative risk for major adverse cardiovascular events (cardiovascular death, MI, ischemic stroke) [68]. On the other hand, a study by O’Connor et al. investigating 73 patients with DM, did not show any correlation between hemoglobin A1 c (HbA1 c) levels and severity of PAD or need for revascularization, suggesting that other factors than DM might play a more important role in PAD course [69].

As it is well known that DM is a strong risk factor for cardiovascular and renal disease, several other agents including angiotensin converting enzyme (ACE) inhibitors, and statins have been considered crucial for the management of PAD and DM. The Heart Outcomes Prevention Evaluation (HOPE) study tested whether ACE inhibitors compared to placebo improve morbidity and mortality in patients being at high risk for cardiovascular adverse events [70]. Almost 3,600 diabetic patients with a history of a cardiovascular adverse event or at least one traditional risk factor for atherosclerosis were included in the study [71]. The HOPE trial was terminated early (i.e. 6 months of follow up) because of consistent superiority of ramipril compared to placebo [71]. Ramipril was associated with 22%, 33% and 37% risk reduction for MI, stroke and cardiovascular mortality respectively [71]. This benefit of ramipril remained statistically significant even after adjusting for the decrease in blood pressure and as such its efficacy was attributed to additional reno- and vasculo-protective effects among diabetics [71]. Moreover, a retrospective analysis of approximately 470 patients with CLI showed that ACE inhibitors and angiotensin II receptor blockers (ARB) were associated with lower mortality and major adverse cardiovascular events in patients with CLI during a follow up period of 3 years [72]. These findings suggest that use of an ACEI/ARB could improve the prognosis and overall survival of CLI patients; however, as major limb adverse events were not prevented by this class of agents, revascularization could be considered the main requirement for successful limb preservation in patients with advanced PAD and DM [11,72].

Statins on the other hand, especially when prescribed in high doses at the time of PAD diagnosis, have been correlated with improved freedom from limb-loss and all-cause mortality [73], with a recent systematic review, including almost 27,000 patients, demonstrating clear benefit of statin use in advanced PAD cases (i.e. CLI) [74]. Hereby, the American Heart Association/American College of Cardiology (AHA/ACC) guidelines recommend that all patients with clinically present atherosclerotic disease should be prescribed high-intensity statins [75]. Interestingly, the observed benefits of statins have been attributed to pleiotropic effects affecting the endothelium, microvascular function and remodeling of the atherosclerotic plaque [76–78]. This was supported by, the Heart Protection Study (HPS; ClinicalTrials.gov Identifier: NCT00461630), a randomized clinical trial including almost 6,800 patients with PAD, which demonstrated that 40 mg simvastatin regardless the baseline cholesterol levels reduced the incidence of first major vascular events (i.e. non-fatal MI, MI or death from CAD, stroke, any type of vascular revascularization) [79]. Therefore, there is strong evidence that statins should be routinely prescribed for all patients with PAD as part of prevention for cardiovascular and major limb adverse events.

In addition to ACE inhibitors and statins, anti-platelet therapy (PLT) has been proven to play a significant role in the PAD management [80]. A major goal of anti-PLT is to reduce the risk of acute limb ischemia caused by thrombosis and to reduce the risk for future major cardiovascular adverse events attributed to atherosclerosis progression [81]. According to TASC II, single anti-PLT is recommended before any endovascular procedure for PAD, continued indefinitely [8], while ACC/AHA guidelines recommend at least 1 month of dual-antiplatelet therapy (DAPT) with aspirin and clopidogrel [82]. Current guidelines for antiplatelet therapy in PAD are summarized in Table 3 [9,34–36]. However, only a few studies guide decision-making regarding the optimal anti-PLT duration after endovascular therapy for PAD [80,82]. Furthermore, although the majority of clinical trials investigating devices for endovascular peripheral interventions mandated 1–3 months DAPT with aspirin and clopidogrel, the additive benefit of DAPT remains unclear [83]. Therefore, the choice of DAPT over single anti-PLT should be individualized and balanced by the potentially higher risk for bleeding with DAPT, while the exact role of DAPT and other anti-thrombotic strategies (e.g. rivaroxaban) in PAD management should be further investigated [84].

Table 3. Current guidelines for antiplatelet therapy in PAD [85].

Patients status	SVS 2015 [36]	AHA/ACC 2016 [35]	ESC 2017 [9]	GVG 2019 (CLI) [34]
Asymptomatic PAD	-	Anti-PLT when ABI values are abnormal	Anti-PLT should be considered (individualized)	-
Symptomatic PAD	Monotherapy with aspirin or clopidogrel	Monotherapy with aspirin or clopidogrel	Monotherapy with aspirin or clopidogrel; clopidogrel should be preferred	Monotherapy with aspirin or clopidogrel for patients with CLI; combined therapy with aspirin and rivaroxaban should be considered
PAD treated with EVT	DAPT with aspirin and clopidogrel for at least one month post-procedurally	DAPT should be considered	DAPT with aspirin and clopidogrel for at least one month followed by lifelong monotherapy	DAPT with aspirin and clopidogrel for at least one month post-procedurally

*EVT: endovascular therapy, PAD: peripheral artery disease, anti-PLT: antiplatelet therapy, DAPT: dual anti-PLT therapy, CLI: critical limb ischemia, SVS: Society of Vascular Surgery, AHA/ACC: American College of Cardiology/American Heart Association Guideline, ESC: European Society of Cardiology, GVG: Global Vascular Guidelines, MACE: major adverse cardiologic event

Interestingly, although therapy with ACE inhibitors, statins and anti-PLT have shown additive benefits in patients with PAD and/or DM [86–88], several population studies have reported undertreatment of PAD patients [89–91]. Furthermore, the ‘A Study Comparing Cardiovascular Effects of Ticagrelor and Clopidogrel in Patients With Peripheral Artery Disease’ (EUCLID; ClinicalTrials.gov Identifier: NCT01732822) study showed that patients with coexistent PAD and DM were at higher risk for cardiovascular and limb ischemic events, despite contemporary risk reducing medical therapies (e.g. blood pressure control, anti-platelets) [68]. This finding is indicative that currently available treatment approaches inadequately address the increased risk for bad prognosis associated with DM [68]. Both cardio- and cerebrovascular complications more commonly occur in patients with combined PAD and DM compared to patients with PAD only [4,25], however the prognosis of those patients with coexistent DM and PAD is also correlated to other comorbidities (e.g. presence of ischemic wound, infection, neuropathy etc.) [92]. Therefore, in order to improve the prognosis of patients with concomitant DM and PAD, increased awareness of the benefits of secondary PAD prevention (e.g. cardiovascular mortality, amputation, MI, stroke etc.) and optimized risk factor modification are needed. Additionally, as there is still controversy regarding the role of glycemic control (i.e. blood glucose level at the time of the index procedure) in PAD management and prevention of cardiovascular mortality, the incidence of amputation could be a reasonable alternative measure in order to investigate the quality of DM care among patients with PAD [93,94]. Thus, further research investigating limb-related outcomes is warranted in order to develop more aggressive treatment algorithms, including optimized medical therapy and revascularization protocols.

Declaration of Interest

E Armstrong is a consultant to Abbott Vascular, Boston Scientific, Cardiovascular Systems Incorporated (CSI), Intact Vascular, Medtronic, Philips, and PQ Bypass. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewer Disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose

Additional information

Funding

This paper was not funded.