Dotter and Judkins, in 1964, described a series of 15 endovascular procedures in nine patients on 11 limbs [1]. Eight of their patients had gangrene, whereas the others had claudication or rest pain. In five of the eight patients, an amputation was averted; an additional amputation was delayed by 3 months. One patient functionally had a hybrid procedure. Their technique involved an “ordinary coil-spring catheter guide of about 0.05 in. OD is passed down the lumen until its tip has traversed the stenosis to reach the lumen beyond.” Notably, some of these vessels were occluded, and some have suggested that this represents the first report of subintimal guidewire placement. Continuing with the procedure, “a tapered, radiopaque, Teflon dilating catheter of approximately 0.1 in. OD is then slipped over the guide and advanced until it, too, has traversed the block, thereby enlarging the preexisting or newly opened lumen.” And, “where desirable and possible, a second dilating catheter of nearly 0.2 in. OD is passed over the first” [1]. Beyond opening the door to a paradigm shift in how these patients can be treated, Dotter and Judkins also were robust in following these patients longitudinally with objective studies—with plethysmography and segmental pressures.

Building upon this, Dorros, Jaff et al. published their series from 1983 to 1996 of 284 limbs in 235 patients with CLI [2]. They had a 92 % success rate in the ability to dilate the tibioperoneal lesions. At 5 years, there was a 91 % limb salvage rate, bypass surgery was performed in 8 %, and survival was only 56 % (lower with more advanced disease) [2]. Notably, the complication rate was low—one procedurally related death, emergency vascular surgery in three patients (arterial access repair, 2; no emergent bypasses , amputation, 1), one infection, one compartment syndrome, one patient required transfusion, and 20 patients (7 %) developed acute renal failure [2].

Natural history studies of critical limb ischemia are limited and have inherent methodological flaws, biased by patient selection. Nonetheless, there are some studies that provide insights into what can be expected from optimal medical therapy. In a series of patients by Marston et al. between 1999 and 2005, they followed 142 patients (169 limbs) with both PAD and a full-thickness ulceration that were not revascularized [3] (Table 34.1). These patients carried the traditional risk factors: mean age of 70 years, 70 % were diabetics, and chronic renal insufficiency (serum creatinine > 2.5 mg/dL) in 28 %. Within 6 months, 19 % required a major amputation and by 1 year 23 %. Foot-sparing amputations were performed in 28 % of limbs. Complete wound healing was seen in 25 % of patients at 6 months and 52 % by 1 year. Limb loss was associated with a lower ABI: with an ABI <0.5, major amputation rate is 28 % (6 months) and 34 % (12 months) compared with 10 % (6 months) and 15 % (12 months) with an ABI > 0.5. The only risk factor that was found to be associated with wound closure was the initial size of the wound [3].

Endovascular treatment of critical limb ischemia assumes that the patient has undergone a vascular assessment. Surprisingly, though, vascular evaluation is not routine prior to amputations. In a study of 17,463 Medicare patients from 2000 to 2010 who underwent nontraumatic amputation, only 68.4 % had some type of arterial evaluation within 2 years of the amputation. In decreasing frequency of testing were: ABI (47.5 %), arterial duplex (38.7 %), invasive angiography (31.1 %), CT-angiography (6.7 %), and MRA (5.6 %) [4]. Frequency of arterial testing varied by the level of amputation: foot, 62.5 %; below the knee, 76.7 %; and above the knee, 62.5 %. The rate of arterial testing increased, though minimally, over the study period: 65.7 % (2002) to 69.2 % (2010 ). The utilization of ABI, arterial duplex, and CT-A all increased over time, whereas invasive angiography remained the same and MRA use declined (Table 34.2). Attention should be made to reduce racial disparities in revascularization rates [5].

Arterial insufficiency represents one component to the multiple etiologies contributing to lower extremity ulcerations [6]. Arterial insufficiency is the etiology of approximately 50 % of foot ulcerations while <25 % of above foot ulcerations [7]. The use of noninvasive studies can help determine the likelihood of ischemic rest pain as well as wound healing in order to guide endovascular therapies. Typically, without local trauma or infection, less flow is required for maintenance of intact integument than is required for wound healing. It is for this reason that it is important to distinguish these cutoffs. Ischemic rest pain may develop at an absolute ankle pressure of ≤50 mmHg and absolute toe pressure of ≤30 mmHg, whereas nonischemic etiologies should be sought in patients with rest pain and pressures above these cutoffs [7]. Review of the literature will reveal other cutoffs, and some have used a cutoff for absolute tissue ischemia with an ankle pressure of ≤35 mmHg in nondiabetic patients and ≤55 mmHg in diabetics [8]. Conversely, ankle pressures ≤70 mmHg or toe pressures ≤50 mmHg may be seen in patients with ischemic tissue loss (Fontaine IV or Rutherford category 5 or 6) [7]. Further confirmation of an ischemic component can be from a severely dampened or flat pulse volume recording (“PVR”) .

Despite efforts to predict which patients require aggressive intervention, it should be remembered that fewer than half of the patients with CLI, requiring a below the knee amputation, had symptoms up to 6 months prior [9]. Coupled with data that shows only about 75 % of below the knee amputations heal by primary or secondary intention, 15 % require an above the knee amputation, and 10 % will die in the perioperative amputation period, highlighting the need for revascularization strategies which minimize physiological stress, patient morbidity, and mortality. Depending on the initial level and type of amputation , reamputation rates for the ipsilateral limb at 1 year in diabetic patients are approximately 25 and 60 % by 5 years; contralateral amputations at 5 years are as high as 50 % [10].

The decision to proceed with endovascular revascularization must be weighed against surgical revascularization, at times, even weighed against conservative and adjunctive therapies. There is a paucity of well-designed trials which compare endovascular to surgical revascularization, let alone various endovascular techniques/devices against each other [11].

In an early series of tibioperoneal angioplasty in 284 limbs, 59 % were found to have some form of inflow treatment prior to the infrapopliteal angioplasty [2]. In a study by Tartari et al., evaluating the efficacy of subintimal angioplasty as the first choice of revascularization in patients with critical limb ischemia, out of 109 limbs, 27 (25 %) of the lesions were isolated to the femoropopliteal segment , 36 (33 %) had combined femoropopliteal and tibial lesions, and 46 (42 %) were isolated to the tibial vessels [12]. Since this was a study of infrainguinal occlusions, iliac lesions were not included in the analysis. Gray et al. demonstrated that in 446 patients with CLI undergoing revascularization, 36 % had isolated tibial disease, 64 % had multilevel disease, and 8 % had both suprainguinal and infrainguinal diseases [13]. In a series of 229 limbs with critical limb ischemia, aortoiliac disease was found in 1.7 %, iliofemoral in 5.2 %, “iliodistal” 3.9 %, femoropopliteal 22.7 %, femorodistal 35.4 %, and below the knee only 31 % [14]. It is estimated that about 13–25 % of patients with CLI will not be amenable to revascularization; ultimately, although lesions may be amenable to endovascular treatment, the clinician must keep in mind that the patient and limb may not ultimately be able to be healed due to comorbidities and other contributing factors [14, 15].

A consecutive series of diabetic patients with critical limb ischemia by Graziani et al. highlights the anatomic complexity which is encountered when an endovascular approach is taken. Of 417 patients, there were 2893 lesions identified. Over half (55 %) of the lesions were occlusions. Only 1 % of the stenosis were in the iliac arteries, whereas 74 % of the below the knee vessels were identified involved. Of these below the knee lesions, 66 % were occlusions (50 % were >10 cm in length). In 55 % there was at least one patent tibial artery and in 28 % of the patients, all three of the tibial vessels were occluded. Not surprisingly, TcPO2 correlated to the extent of vascular involvement [16]. TASC II uses a TcPO2 threshold of ≤30 mmHg for diagnosis of CLI and if <20 mmHg revascularization is often necessary for wound healing; similarly, skin perfusion pressure (SPP) using laser Doppler of <30 mmHg is frequently seen in patients with CLI [17].

Faglia et al., in a series of nearly 1000 diabetic limbs, determined the level of disease contributing to CLI. Lesions of >50 % were found in the iliac artery, femoral artery, or popliteal artery in 6.7 %, entirely in the below the knee arteries (infrapopliteal ) in 31.8 %, and in both the fem-pop and infrapopliteal segments in 61.4 % of patients [18].

Angiosome-directed interventions will be reviewed in a separate chapter. However, the ability to achieve revascularization of the “appropriate” angiosome artery can be technically difficult, if not impossible, endovascularly. Outcomes, although seemingly should be improved with this approach, often times oversimplify the relationship between the tibial vessel above the ankle and the status of the plantar vessels and plantar arch. Acin et al. analyzed the outcomes of infrapopliteal artery interventions relative to the number of tibial vessels attempted, number of patent tibial vessels, and if direct revascularization (angiosome appropriate) or indirect revascularization (ulcer supplied by collaterals) [19]. They found that ulcer healing at 1 year was equivalent between the direct (66.0 %) and indirect (68.0 %) revascularization strategies as was limb salvage at 2 years (88.9 % vs 84.8 %, respectively) [19]. Indirect revascularization without collaterals resulted in a 7.1 % 1 year ulcer healing rate, though limb salvage at 2 years in this group was 59.0 %. Amputation-free survival at 24 months was statistically equivalent between the groups (direct revascularization, 67.5 %; indirect revascularization through collaterals, 73.3 %; and 61.9 % indirect revascularization without collaterals) [19]. Amputation-free survival at 2 years was lower in the group in which there were no patent (<30 % residual stenosis) tibial vessels (43.8 %), but statistically equivalent between runoff “1” and runoff “>1” groups (64.3 % vs 76.6 %) [19]. Further contributing to the confusion surrounding whether angiosome-guided revascularization enhances outcomes is data from Kawarada et al. They evaluated the skin perfusion pressure after revascularization of either the anterior tibial or posterior tibial artery. They found that comparing the dorsal side to the plantar side, anterior tibial artery revascularization, 64 and 58 %, demonstrated higher dorsal post-intervention skin perfusion pressure and change in skin perfusion pressure; posterior tibial artery revascularization resulted in higher plantar skin perfusion pressures in 47 and 40 % had a higher post change in skin perfusion pressure. They concluded that “single tibial artery revascularization , whether of the ATA or PTA, yielded comparable improvements in microcirculation of the dorsal and plantar foot. Approximately half of the feet revascularized had a change in microcirculation that was not consistent with the 2D angiosome theory ” [20].

The TransAtlantic Inter-Society Consensus (TASC) II has developed a scheme by which lesions can by approached from an anatomic perspective. Lesions are classified A–D. Type A lesions are preferentially treated with an endovascular approach, whereas type D lesions are best suited for surgical revascularization [7]. These recommendations were published in 2007, however, and there have been significant advancements in endovascular techniques which have allowed for the greater utilization of endovascular procedures for lesions that otherwise would have been reserved for vascular surgery. The Society of Cardiac Angiography and Interventions has published expert consensus statements which address appropriate endovascular treatment of lesions based upon anatomy and symptom status [21, 22].

The goal of this chapter is not to discuss particular methodologies or techniques but rather to provide an overview of endovascular principles and particularly to discuss outcomes with an endovascular strategy. For an exhaustive review on endovascular techniques, please refer to our separate textbook, Endovascular Interventions: A Case Based Approach [23].

The Society for Cardiovascular Angiography and Interventions (SCAI) published expert consensus statements on appropriate revascularization of lower extremity arterial disease [21, 22]. Table 34.3 details how to approach a patient with infrapopliteal disease. These recommendations follow a rational and logical approach to the clinical scenario based upon the patients symptoms classification paired to the artery anatomy.

Arterial access is dictated by a number of factors, including lesion location, body habitus, and local tissue issues (such as hernias, scar tissue, infection, etc.). Furthermore, it is not uncommon for catheter or wire length to limit approaches. Although radial access can be used for some aortoiliac lesions, it is difficult for even a 90 cm sheath to engage the iliac arteries in some patients from a radial approach. Similarly, if a contralateral common femoral approach is used, equipment may not reach the contralateral dorsalis pedis artery . As such, the interventionalist must be cognizant of catheter limitations, access vessel diameters that may limit equipment being used, as well as hemostasis issues. A variety of techniques are available—even high anterior tibial access [24]. Antegrade crossing for tibial intervention is unsuccessful in up to 20 % of patients. In a study of 51 failed antegrade crossing attempts, a transcollateral approach (11.8 %) or pedal access (88.2 %) was utilized resulting in an 86.3 % success rate [25]. Direct digital arterial access has also been utilized for revascularization.

There are a number of techniques described for lesion crossing. For stenosis, true lumen crossing is relatively straightforward, particularly with road mapping techniques. Occlusions however can be crossed with an attempt to remain in the true lumen. The longer the occlusion or more calcified the lesion, remaining true lumen can be difficult.

Bolia et al. first introduced subintimal angioplasty (SIA or percutaneous intentional extraluminal recanalization—PIER ) in 1989. Their report of the inadvertent creation of a subintimal channel in the popliteal artery and subsequent angioplasty introduced a new technique which allows an alternative than true lumen lesion crossing [26]. True lumen crossing in long-segment chronic total occlusions (CTOs) can be difficult and identification of the vessel course can be challenging. In a study by Hynes et al. comparing the “pre-SIA” period and the period after the introduction of SIA, “post-SIA” over 15 years, the limb salvage rate increased from 42 to 70 %, in the pre to post SIA period. The post revascularization limb salvage rate increased from 72 to 86 % and the 30-day morbidity, mortality, and length of stay were shorter [27]. Zhu et al. evaluated the efficacy of subintimal angioplasty in the dorsalis pedis or plantar arteries in diabetic patients with CLI [28]. They were successful in 83.3 % of cases with limb salvage rate of 94.6 % [28]. Many of the techniques used for tibial artery occlusions were originally developed for the treatment of coronary artery CTO. The controlled antegrade and retrograde tracking and dissection (CART) /reverse CART technique is one such method, originally described by Surmely in 2006 [29]. When balloons are simultaneously inflated/deflated, the technique is referred to as “confluent balloon technique” [30]. Similarly, the parallel wire technique in which after the first guidewire enters the false lumen, a second wire is used with the first wire as a guide, was first introduced by Japanese coronary CTO experts [31].

There are, however, no randomized controlled trials comparing SIA to other lesion crossing techniques and vascular endpoints are diverse across studies. In an analysis of 11 studies, involving approximately 1400 limbs, Brennan found that technical success in patients with CLI was 87.7 %. One-year primary patency was 55.8 % (four studies) and 1-year limb salvage rate was 91.1 % (six studies) [32]. This data again highlights the disparity between primary patency (relatively low) and the limb salvage rates (relatively high). Another systematic review of SIA demonstrated that technical success is higher in femoral or femoropopliteal arteries compared to crural vessels [33]. Complication rates were reported at about 15 % and ranged from puncture site hematoma, perforation, or distal embolization with a range of 2–20 % across studies. One-year primary and assisted primary patency rates are about 50 % with limb salvages rates about 80–90 % [33]. Vraux et al. demonstrated a 77 % success rate of SIA for tibial vessels with 1-year primary patency rates of 46 and 87 % 1-year limb salvage rates [34].

The utilization of a transcollateral approach has allowed increased success rates where antegrade crossing has failed. The ability to access the target vessel via collaterals and treat the occlusion has the advantage of one access point (as opposed to a separate pedal access). Current small profile balloons will generally track well and allow for dilation in this retrograde manner. After angioplasty, the vessel can either be rewired from an antegrade approach or the retrograde wire externalized, maintaining lumen access [25].

Complications of endovascular interventions, including contrast induced nephropathy, must be weighed against the potential benefit to the patient. Embolization from peripheral interventions can lead to additional procedures, prolonged hospital stays, convert a stable patient into acute limb ischemia, and in some cases loss of limb or death. Distal embolization is seen in 1–100 % of peripheral interventions. Rates are dependent upon how embolic debris is detected—“hits” by ultrasound, capture with filters, or clinically relevant embolization. When an embolic protection device is used for endovascular interventions, Mendes et al. found macroscopic debris in almost 70 % of cases [35]. Embolization leads to increased reintervention (20 % vs 3 %) and amputation rates (11 % vs 3 %) [35]. Embolization rates are likely related to lesion complexity (occlusions), and atherectomy devices appear to have a higher rate of embolization (Fig. 34.1).

All percutaneous vascular interventions carry the risk of the development of an access-related arteriovenous fistula (0.4 %) or pseudoaneurysm . Interventional procedures, presumably due to the use of anticoagulation, increase the risk from approximately 0.05–2.0 % for diagnostic cases to 2–6 % for interventional procedures [36]. Other risk factors include increasing sheath size, female gender, and inadequate compression post sheath removal [23]. Spontaneous resolution of small pseudoaneurysms <2–3 cm in patients not on anticoagulation can be expected in almost 90 % of patients. Pseudoaneurysm morphology guides treatment strategies. Generally, if the pseudoaneurysm is small, without an identifiable neck, has a concomitant arteriovenous fistula, or is in very close proximity to other vessels, then ultrasound-guided compression (or even non-ultrasound-guided compression) should be attempted first (approximately ≥75 % success rate); otherwise, the pseudoaneurysm is treated with ultrasound-guided thrombin injection (≥90–95 % success rate).

Dosluoglu et al. reported on limb salvage and survival after adoption of an endovascular first approach to critical limb ischemia [37]. The 30-day mortality trend was lower in the endovascular vs open group (2.8 % vs 6.0 %, p = 0.079). Five-year limb salvage rates were equivalent to 78 %, amputation-free survival in the endovascular group was 30 % vs 39 % in the surgical group, and overall survival was equivalent to 36 % vs 46 % [37]. Although primary patency rates were equivalent at 5 years (50 % vs 48 %), the assisted primary patency rates (70 % vs 59 %) and the secondary patency rates (73 % vs 64 %) favored the endovascular first approach [37].

A retrospective analysis of 1053 patients with critical limb ischemia by Soga et al. compared treatment with either bypass surgery or endovascular therapy first for limb revascularization [38]. At 3 years, the groups were equivalent in limb salvage (85.4 % vs 88.7 %), amputation-free survival (62.1 % vs 60.5 %), and overall survival rates (69.2 % vs 65.8 %). A matched pair analysis confirmed the equivalency between the two groups [38].

Jones et al. performed a meta-analysis of studies with patients with critical limb ischemia comparing the effectiveness of endovascular versus surgical revascularization [39]. Overall, 23 studies (only one randomized control trial) were used in the analysis. There was no difference in overall mortality between 1 and 2 years or after 3 years. The rates of lower extremity amputation were equivalent at <2 years, at 2–3 years, and at >5 years. Amputation-free survival was equivalent between the groups at these time points, as well. Endovascular treatment had higher primary patency and secondary patency at 1 year and at 2–3 years [39].

In an analysis of 202 patients involving 229 limbs with critical limb ischemia, May et al. evaluated the efficacy of an endovascular first approach [14]. Patients preferentially underwent an endovascular approach unless there were contraindications to angiography (11 patients), anatomy not amenable to percutaneous intervention (20 patients), or those that failed an endovascular approach (16 patients). Ultimately, 198 limbs underwent endovascular revascularization. One hundred and forty-four of these required no further intervention (72.7 % of those underwent endovascular intervention), 38 (19.2 %) required a secondary endovascular intervention, and 16 (8.1 %) required bypass surgery after an initial endovascular approach, but with continued impaired perfusion and wound nonhealing (mean 37±58 days after the initial endovascular intervention) [14]. The mortality rates at 30 days, 12 months, and 24 months were 5.2 %, 20 %, and 27 %, respectively. However, the cumulative limb salvage rates and amputation-free survival at 1 year were 78 and 75.5 % and, at 2 years, 74 and 57.6 % [14]. Interestingly, in this study, the investigators opted not to report the vessel patency, citing reluctance of asymptomatic patients to undergo further testing, but also “limb preservation is the ultimate condition that matters most to the patients.”

Another study, by Garg et al., evaluated the difference between endovascular first approach and surgical bypass [40]. Their findings support an equivalent outcome between the two strategies in regard to 5-year survival and limb salvage, but more secondary procedures after an initial open strategy. Patients with end-stage renal disease, below the knee interventions, or gangrene had a lower amputation-free survival rate [40].