Abstract
Keywords
Introduction
The goal of endovascular limb salvage is to obtain direct flow to the foot lesion based on the angiosome concept 1 to promote healing and avoid amputation. However, there are challenging situations in which conventional revascularization techniques still fail to have a tangible positive impact on amputation rates, such as in dialysis patients with long, severely calcified below-the-knee (BTK) vessels and heel ulcers. This patient subset has a high failure rate for intraluminal or subintimal recanalization, 2 as well as high amputation rates despite excellent operator skills and improvement in endovascular techniques. Thus, further treatment strategies must be developed to reestablish direct flow and ameliorate clinical outcomes.
Although surgical deep vein arterialization has been performed for decades to salvage critically ischemic limbs with no reconstruction options,3,4 Kum et al 4 recently published encouraging results from the first pilot study of percutaneous deep vein arterialization using the LimFlow device. Herein is presented another technique for complete endovascular distal plantar vein arterialization (DPVA) in no-option critical limb ischemia (CLI) patients (Rutherford category 6) with severely calcified occlusion of the posterior tibial artery (PTA) and poor plantar artery runoff.
Technique
The technique is demonstrated in a patient with occluded posterior tibial, medial plantar, and lateral plantar arteries (Figure 1A-C). Preoperative dual antiplatelet therapy with 100 mg/d of acetylsalicylic acid and 75 mg/d of clopidogrel is started at least 72 hours before the procedure. After administration of local anesthesia (10 mL of 2% lidocaine), the ipsilateral common femoral artery is punctured in antegrade fashion using the Seldinger technique. A 6-F, 25-cm-long introducer sheath (Radifocus Introducer II, Terumo, Tokyo, Japan) is inserted, and heparin is administered to maintain the activated clotting time >250 seconds.
(A, B, C) Preprocedure angiography showed patency of the anterior tibial and peroneal arteries but absent posterior tibial (PTA), medial plantar, and lateral plantar arteries (circle in the anteroposterior view in C). (D) After an unsuccessful intraluminal attempt, subintimal recanalization of the PTA was attempted by advancing the 0.014-inch guidewire with a narrow U-shaped loop (arrow). (E) The loop (arrow) was advanced as distally as possible in the medial plantar artery followed at a short distance by the Armada XT 2-×20-mm balloon catheter (circle). (F) In this case (patient 1), a fortuitous entry in the medial plantar vein was obtained with the tip of the Terumo Glidewire Advantage, oriented medially to the vessel wall calcifications (circle) in the anteroposterior view. (G) The guidewire was advanced in the lumen of the medial plantar vein beyond the midfoot (arrow); (H) the entry into the vein lumen was confirmed by gentle injection of contrast through the balloon catheter with a 2.5-mL syringe (arrows). After 2.5-×150-mm balloon catheter dilation, (I) postprocedure angiography showed patency of the PTA. (J) Good retrograde filling of the small distal venous channels was seen in the nonhealing wound area (circle).
After failure of BTK and below-the-ankle intraluminal recanalization attempts, a subintimal approach through the PTA is begun using a 0.014-inch, 190- or 300-cm-long guidewire (Glidewire Advantage; Terumo, Tokyo, Japan) supported by a 2-×20-mm, low-profile (0.017-inch), over-the-wire balloon catheter (Armada XT; Abbott Vascular, Redwood City, CA, USA) positioned a short distance behind the narrow U-shaped loop in the guidewire (Figure 1, D and E). If passage is difficult, the wire is withdrawn and readvanced to create another subintimal channel.
Typically, due to the heavy calcification in the distal tortuous segment of the PTA (and/or in the lateral or medial plantar artery) no reentry to the arterial true lumen can be achieved despite multiple attempts. However, an entry in the distal lateral or medial plantar vein from a subintimal channel in the plantar artery (Figure 1F) can be intentionally pursued as a bailout technique, pointing the tip of the guidewire opposite to the arterial wall calcifications. The best angiographic projection to perform this maneuver is generally the anteroposterior (AP) view because the plantar veins are usually doubled, and plantar arteries lie between this doubled segment. 5
In the AP projection, with a slight craniocaudal orientation of the C-arm, the guidewire is advanced toward the external side of the artery, that is, medially when it is in a medial subintimal channel of the plantar artery (as shown in Figure 1F) or laterally when it is in a lateral subintimal space of the artery. When the artery wall is breached, vein access is confirmed by gentle injection of contrast through the balloon catheter with a 2.5-mL syringe (Figure 1H). In cases of unsuccessful entry in the vein lumen, the guidewire and balloon catheter are withdrawn, another medial or lateral subintimal channel is sought, and the maneuver is repeated. An average of 15 minutes of fluoroscopy is generally allowed to access a vein lumen.
Once the guidewire is advanced in the distal lateral or medial plantar vein and a plantar arteriovenous fistula (AVF) has been created (Figure 1H), angioplasty of the arteriovenous anastomosis and the occluded PTA segment is performed by inflating 0.014-inch platform balloon catheters with diameters and lengths ranging between 2.5 to 3 mm and 80 to 210 mm, respectively. In this case, PTA patency and good retrograde filling of the small distal venous channels in the nonhealing wound area were obtained (Figure 1, I and J).
Clinical Experience
To October 2016, 9 consecutive diabetic, end-stage renal disease (ESRD) patients (mean age 69 years; 5 men) with no-option CLI underwent endovascular limb salvage procedures. All patients presented with nonhealing ulcers (5 heel, 3 plantar, and 1 transmetatarsal) of varying severity according to the WIfI classification system 6 (Table 1) and long PTA chronic total occlusion with poor or absent plantar artery runoff. Four patients had “desert foot,” angiographically described as the lack of both plantar arteries, the plantar arch, and the dorsalis pedis and lateral tarsal arteries. Six patients had no perfused plantar arteries.
Baseline and Procedure Characteristics.
Abbreviations: LPV, lateral plantar vein; MPV, medial plantar vein; TMT, transmetatarsal after previous amputation; WIfI, wound, ischemia, foot infection.
In all cases, the plantar arteries were accessed with the Armada XT low-profile balloon catheter and the 0.014-inch Glidewire Advantage. An AVF was created with the surrounding plantar vein in 7 of 9 patients, creating direct blood flow to the wound. A detailed description of the level of venous entry in the lateral or medial plantar vein before or beyond the midfoot can be found in Table 1. As no valves were observed in the runoff foot veins, no adjunctive valve lysis or disruption was required. Postprocedure angiographic controls showed good retrograde filling of the small distal venous channels in the nonhealing wound area (Figures 1 and 2).
(A) Preprocedure angiography showed patency of the anterior tibial artery and occlusion of the peroneal artery and posterior tibial artery (PTA) in patient 3. (B) Very small medial and lateral plantar arteries were recognizable. (C) Subintimal recanalization of the PTA was attempted, advancing the 0.014-inch guidewire with a narrow U-shaped loop (arrows in B and C) supported by the low-profile balloon catheter. After entry into the lateral plantar vein through the subintimal space of the lateral plantar artery, the arteriovenous fistula and the PTA were dilated with a 2.5-×120-mm balloon catheter. (D) Postprocedure angiography showed patency of the PTA and good retrograde filling of the small distal venous channels in the nonhealing wound area (circle).
Mean fluoroscopy times averaged ~65 minutes; 3 cases were successfully completed with <30 minutes of total fluoroscopy. Some arterial perforation occurred without any clinical significance. During the first 4 days after the procedure, a transitory mild edema and purplish discoloration of the treated foot were observed in 6 of 7 patients, which disappeared within the first week. No major complications were observed. The mean transcutaneous oximetry (TcPO2) readings before and after revascularization were 7.3±2.2 and 37.1±17 mm Hg, respectively. The successfully treated patients were discharged on dual antiplatelet therapy. The 2 patients in whom an AVF was not created underwent BTK amputation 1 week after the unsuccessful procedure.
Imaging surveillance to determine the patency of the AVF was not performed but was indirectly evaluated by clinical examination and oximetry in follow-up. In the 7 successfully treated patients, the mean TcPO2 at 1 month was 30±17 mm Hg. Six ulcers healed over an average of 21±4 weeks. The seventh patient underwent BTK amputation 1 week after the procedure due to impairment of the heel ulcer. One subject had a relapse of lower limb ischemia 2 months after revascularization. Occlusion of the PTA and the AVF prompted a second but unsuccessful endovascular intervention followed by BTK amputation. No patients died during the 6-month mean follow-up.
Discussion
Severity of CLI is related to the degree of BTK vessel calcification and target vessel diameter. Heavily calcified vessels with a diameter <3.0 mm and length >300 mm, especially in the PTA with no BTA runoff, are associated with high rates of major amputation after limb salvage procedures.7,8 Furthermore, clinical efficacy of BTK angioplasty in ESRD patients is limited due to the severely diseased foot arteries; hemodynamic improvement is obtained in only half of the patients. 9 Additionally, in diabetic patients, dialysis is a negative predictor of BTK revascularization success and a positive predictor of major amputation. Therefore, recanalization of heavily calcified BTK vessel occlusions in dialysis patients can be very challenging. 2 Moreover, in case of nonhealing heel ulcers, a BTK amputation is mandatory, leading to a poorer prognosis. 10
In our opinion, because of the natural tortuosity of the PTA and its typically small diameter, a long heavily calcified occlusion may be extremely challenging to recanalize, which could explain the poor prognosis of heel ulcers. While linear “railroad track” calcifications can ease the guidewire’s advancement through the artery by providing a natural “roadmap,” coarse and patchy calcifications can hinder not only intraluminal progress but also intraluminal reentry from a subintimal space. Conversely, patchy calcifications might steer the guidewire’s advancement toward the external layer of the arterial wall, which could be perforated to access the adjacent vein lumen.
Our first case was the impetus for development of a completely percutaneous DPVA technique in which the PTA is connected distally to the medial or lateral plantar vein through an AV subintimal channel. The creation of the AVF in this patient was accidental, and an endovascular revision of the fistula was initially considered. However, on the basis of clinical findings and TcPO2 measurements, we soon realized the potential of this technique as the ultima ratio after the failure of conventional approaches.
Distal arterialization of the venous bed to provide an oxygen source to ischemic peripheral tissues was conceptualized more than a century ago. 11 There are at least 2 mechanisms of action that could explain why venous arterialization might improve peripheral wound healing: neovascularization triggered by angiogenic growth factors or reversed perfusion of arterioles through the rich capillary network of the foot, resulting in increased oxygenation to the ischemic tissues. 12 Perfusion angiography, a technique we unfortunately do not use for CLI in our clinical practice, would have provided valuable information about postprocedure blood flow directed to the wound area.
Many surgical experiences have been described in the literature.3,13–15 In 1975, Lengua 13 introduced the concept of a more distal anastomosis with good outcome. In the following years, other vascular surgeons emphasized 2 crucial points to improve results: a more distal location of the AV anastomosis and the disruption of the valves in veins distal to the site of the anastomosis, including the valves in foot veins.14,15 We did not need to disrupt valves because a terminal anastomosis in the distal lateral plantar vein reduces the probability of encountering valves that might obstruct blood flow, as suggested by Binns and Pho. 5 Moreover, in case of valves in the distal venous bed below the AVF, the high pressure of the bloodstream itself could be strong enough to overcome the valvular barrier, although this concept remains controversial. 16
The first use of a totally percutaneous deep venous arterialization technique was recently described by Kum et al, 4 who employed dedicated arterial and venous ultrasonic catheters to create the AVF. The catheters are advanced respectively in an antegrade and retrograde fashion and then aligned at a level in the proximal tibial artery and vein selected on the basis of the angiosome/venosome concept. A 0.014-inch guidewire is subsequently driven through the arterial inflow into the retrograde venous sheath by means of a dedicated entry device. Then the AVF is predilated with a balloon catheter and a valvulotome is used to disrupt the valves in order to allow unhindered proximal-to-distal blood flow. Finally, a self-expanding covered stent is deployed at the level of the AVF with the proximal extremity in the artery and the distal extremity in the arterialized vein; an endoconduit is created to the ankle level with multiple covered stents. This technique was performed in 7 no-option CLI patients with an admirable 100% technical success and complete wound healing in 4 patients at 6-month follow-up.
Though the technique that we propose herein is less codified, it has the advantage of being less expensive than the technique proposed by Kum et al, 4 with similar clinical results and no permanent implants. Another advantage of our technique is that the distal runoff created by arterialization of the venous bed can sustain the blood flow toward arteries that branch above the AVF in the PTA recanalized segment. Although AV shunting could theoretically increase the blood return toward the heart, the fistula is very distal, and the overall magnitude of the centripetal blood flow is minimal. This could be a likely explanation of why no patients have developed congestive heart failure symptoms.
Insofar as technique herein described, when reentry in the true lumen of the plantar arteries from a subintimal space is not obtained, an entry in the surrounding plantar veins is attempted according to the venosome concept, which generally mirrors the angiosome theory. Selection of the target vein follows the selection of the target artery 1 ; in the case of heel lesions, the target vessel is the medial calcaneal artery, which generally branches at the origin of the lateral plantar artery or at the distal inframalleolar segment of the PTA. For plantar ulcers the target vessel is the medial or lateral plantar artery depending on the location of the ulcer.
Arterial calcifications, very common in this subset of patients, can be usefully exploited to direct the guidewire in the correct direction. However, this is a “blind technique” that does not include phlebography and veins are not opacified, thus their lumen can be pursued only on the basis of anatomical references. Therefore, multiple attempts performed on different levels of the plantar arteries may be necessary to enter into the surrounding vein without fear of damaging a free-flow vessel.
Conclusion
Although further investigations enrolling larger patient populations are required to validate the clinical impact of PDVA, it may represent a promising technique to increase recanalization rates in limb salvage procedures, especially in diabetic dialyzed patients with extremely calcified BTK vessels.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.