Predictors of failure and success of tibial interventions for critical limb ischemia

Abstract

Objective

The efficacy of tibial artery endovascular intervention (TAEI) for critical limb ischemia (CLI) and particularly for wound healing is not fully defined. The purpose of this study is to determine predictors of failure and success for TAEI in the setting of CLI.

Methods

All TAEI for tissue loss or rest pain (Rutherford classes 4, 5, and 6) from 2004 to 2008 were retrospectively reviewed. Clinical outcomes and patency rates were analyzed by multivariable Cox proportional hazards regression and life table analysis.

Results

One hundred twenty-three limbs in 111 patients (62% male, mean age 74) were treated. Sixty-seven percent of patients were diabetics, 55% had renal insufficiency, and 21% required hemodialysis. One hundred two limbs (83%) exhibited tissue loss; all others had ischemic rest pain. All patients underwent tibial angioplasty (PTA). Tibial excimer laser atherectomy was performed in 14% of the patients. Interventions were performed on multiple tibial vessels in 20% of limbs. Isolated tibial procedures were performed on 50 limbs (41%), while 73 patients had concurrent ipsilateral superficial femoral artery or popliteal interventions. The mean distal popliteal and tibial runoff score improved from 11.8 ± 3.6 to 6.7 ± 1.6 (P < .001), and the mean ankle-brachial index increased from 0.61 ± 0.26 to 0.85 ± 0.22 (P < .001). Surgical bypass was required in seven patients (6%). The mean follow up was 6.8 ± 6.6 months, while the 1-year primary, primary-assisted, and secondary patency rates were 33%, 50%, and 56% respectively. Limb salvage rate at 1 year was 75%. Factors found to be associated with impaired limb salvage included renal insufficiency (hazard ratio [HR] = 5.7; P = .03) and the need for pedal intervention (HR = 13.75; P = .04). TAEI in an isolated peroneal artery (odds ratio = 7.80; P = .01) was associated with impaired wound healing, whereas multilevel intervention (HR = 2.1; P = .009) and tibial laser atherectomy (HR = 3.1; P = .01) were predictors of wound healing. In patients with tissue loss, 41% achieved complete closure (mean time to healing, 10.7 ± 7.4 months), and 39% exhibited partial wound healing (mean follow up, 4.4 ± 4.8 months) at last follow up. Diabetes, smoking, statin therapy, and revascularization of >1 tibial vessel had no impact on limb salvage or wound healing. Re-intervention rate was 50% at 1 year.

Conclusions

TAEI is an effective treatment for CLI with acceptable limb salvage and wound healing rates, but requires a high rate of reintervention. Patients with renal failure, pedal disease, or isolated peroneal runoff have poor outcomes with TAEI and should be considered for surgical bypass. (J Vasc Surg 2010;52:834-42.)

Although patients with peripheral artery disease presenting with critical limb ischemia (CLI; rest pain and tissue loss, Rutherford classes 4, 5, 6) have been traditionally treated with surgical bypass, advances in endovascular techniques, including subintimal angioplasty, as well as advances in device technology, have allowed for the successful treatment of more complex patterns of disease. Multiple series have reported on the successful treatment of limb threatening ischemia with endovascular interventions at the femoral and popliteal levels.^1-3

The recently published Trans Atlantic Inter-Societal Consensus document (TASC II) promotes endovascular techniques including angioplasty and stenting as first-line therapy for symptomatic femoropopliteal stenotic or occlusive lesions up to 10 cm in length.⁴ However, the recommendations for infra-popliteal disease are not as clear because of limited data on the efficacy of tibial artery endovascular intervention (TAEI) for CLI in terms of wound healing and limb salvage. There are, however, several recent reports of acceptable patency and limb salvage rates with infrapopliteal interventions for the treatment of CLI.^5-7 This study sought to define predictors of success and failure for TAEI in the treatment of critical limb ischemia and, in particular, the ability of TAEI to achieve wound healing and alleviate rest pain.

METHODS

Patient population

Patients who had undergone infra-inguinal endovascular revascularization, which included TAEI between September 2004 and October 2008 were retrospectively identified from a prospectively maintained database. Indications for treatment included rest pain (Rutherford class 4) and/or tissue loss (Rutherford class 5 and 6). Patients who presented with acute ischemia or who were treated for claudication were excluded. Patient characteristics, co-morbidities, intervention sites, and complications were recorded. Clinical outcomes, including primary patency, primary-assisted patency, secondary patency, limb salvage, and wound healing rates were determined, and preprocedure angiograms were reviewed to assess baseline and postprocedural distal popliteal and tibial runoff.

Endovascular approach

It is the author's primary approach to attempt an endovascular intervention first in all patients, regardless of anatomic limitations. Patients presenting with CLI who cannot be treated by endovascular intervention are offered surgical bypass or primary amputation, depending on the clinical picture. Some of the patients treated in this series, however, were referred by other surgeons for an attempt at endovascular intervention because they were either medically high-risk for surgery or had been determined to have no distal target for a bypass, therefore representing a medical or anatomical high-risk group. Some patients had a failed attempt at endovascular recanalization, and these patients were not included in this analysis.

All procedures were performed by vascular surgeons using either a fixed-imaging hybrid operating room or in the interventional angiography suite. All procedures were performed under local anesthesia with moderate conscious sedation. Contralateral retrograde common femoral access was most commonly performed (78.5%), whereas ante-grade ipsilateral access or trans-brachial access was selectively used. Interventions were performed after systemic heparinization (100 U/Kg).

All interventions were made with the intention of establishing in-line flow into the foot. Decisions on which tibial vessels to treat were made primarily based on the angiographic appearance. If a single revascularized vessel did not appear to supply the area of tissue loss with in line flow, then attempts at intervening on more than one tibial vessel were made with the goal of establishing >1 vessel runoff to the foot. If a single revascularized tibial vessel appeared to supply the area of the wound, then further attempts at tibial vessel revascularization were guided by the ease of such an extra step based on lesion anatomy and by the decision of the operating surgeon.

For complete occlusions, the lesions were crossed in a subintimal plane, and re-entry was confirmed with contrast injection prior to intervention. The wire primarily used is the 0.35 floppy glide wire (Terumo Interventional Systems, Somerset, NJ) along with the quickcross catheter (Spectra-netics Corporation, Colorado Springs, Colo). While a subintimal plane is often used primarily for femoropopliteal lesions, an intraluminal plane is attempted first in tibial lesions. As such, a subintimal plane was not commonly used in tibial total occlusions; however, for short occlusions, it is difficult to exactly determine whether any element of subintimal passage had occurred.

Balloon diameter was selected based on the angio-graphic measurements of the nondiseased arterial segment proximal and distal to the lesion. Self-expanding stents were implanted for femoropopliteal disease following balloon angioplasty and generally were placed for the treatment of residual stenosis (>30%) or flow limiting dissection, at the discretion of the operating surgeon. Stenting of the origin of the superficial femoral artery, the retro and infrageniculate popliteal artery, and the tibial vessels was generally avoided. In addition to angioplasty, initial plaque debulking of tibial lesions was performed with excimer laser atherectomy (Spectranetics Corporation) at the discretion of the operating surgeon. Excimer laser atherectomy was used in both long tibial stenoses and total occlusions. This was done especially for orifice or bifurcation lesions and to cross total occlusions in cases of failure of wire passage, using the 0.9 probe and the step-by-step technique. Interventions were performed with the intention to treat all levels of disease in an attempt to obtain in-line flow to the foot.

All patients undergoing endovascular intervention were treated with an antiplatelet agent, either clopidogrel or aspirin, unless there was a clear contraindication.

Definitions and classifications

Primary patency was defined as the absence of restenosis, occlusion, or re-intervention in the treated arterial segment. The primary patency for any intervention ended when there was clear evidence of occlusion on imaging, if there was a need for repeat endovascular intervention, surgical bypass, or amputation. The need for re-intervention was based on either a return of the patients symptoms with abnormal noninvasive testing (ankle brachial index [ABI] decrease >0.15, dampened pulse volume recordings, or evidence of stenosis by duplex ultrasound scan); duplex scan evidence of recurrent disease alone or worsening of any patient's wound. The duplex ultrasound criteria utilized for the detection of a hemodynamically significant restenosis in an arterial segment previously treated with angioplasty were a peak systolic velocity (PSV) of >300 cm/sec or a velocity ratio (Vr) >3.0. The criteria utilized for the detection of a significant (>80%) in-stent stenosis were PSV >275 cm/ sec and a Vr >3.5.⁸ Noninvasive vascular laboratory surveillance of the treated segments and tibial runoff was performed at 1 month, 3 months, and 6 months post-procedure. Patients were then evaluated at 6-month intervals thereafter. When duplex evaluation was limited due to artifact or body habitus, the toe pressures and quality of PVR tracings were used as surrogates of adequacy of per-fusion along with clinical evaluation of wounds and symptoms. Patients who were found to have asymptomatic duplex evidence of tibial restenosis on follow up were observed and only underwent repeat angiogram and attempted re-intervention for recurrent symptoms.

Assisted primary patency was achieved via secondary endovascular interventions to treat restenoses involving the originally treated arterial segment. Additional procedures to treat lesions proximal or distal to the initially treated segment were also considered secondary interventions to achieve primary-assisted patency. Secondary patency was achieved utilizing secondary endoluminal procedures, which involved recanalizing occluded arterial segments. In those patients who went on to surgical bypass, any patency of the initial endovascular intervention was considered lost at the time of decision for bypass grafting.

Preprocedure popliteal and tibial runoff score were calculated according to a modification of the Society for Vascular Surgery (SVS) criteria, as previously published by Davies et al.⁹ The pre- and postintervention angiographic images for all patients included in the study were reviewed in order to appropriately calculate the distal popliteal and tibial runoff score. The three tibial vessels and the distal popliteal artery were each assigned a score based on the degree of disease. The individual vessel received a score of 0 for <20% stenosis, 1 for 21% to 49% stenosis, 2 for 50% to 99% stenosis, 2.5 for occlusion of less than half the length of the vessel, and 3 for occlusion for more than half the length of the vessel. Each tibial contributes 0 to 3 points to the total score, and the popliteal score is multiplied by 3 with 1 point added to give additional weight to popliteal disease. Therefore, a higher score indicates more severe disease, and there is a maximum score of 19. Three runoff score groups were identified: <5 (Good), 5 to 10 (Compromised), and >10 (Poor).

The wound care regimen was at the discretion of the treating surgeon, who followed all wounds post-intervention. In general, wound care with debridement was performed at each outpatient visit as indicated, and wound dimensional measurements were recorded. Necrotic tissue was sharply debrided, and enzymatic debridement was only offered to patients with evidence of minor fibrinous exu-date on follow up or who could not tolerate office based debridement. If the wound appeared infected with signs of inflammation, systemic antibiotics were administered as well as topical antibiotic therapy. Once the infection was cleared, routine wound care was resumed.

Wound healing was considered poor when the wounds were noted to be failing to improve by 4 weeks from revascularization, if they were noted to be worsening/enlarging at the last documented follow up, or if they progressed to major amputation. In general, patients noted to have worsening of their wounds but with salvageable extremities were treated with repeat revascularization.

Statistical analysis

An independent statistician performed all advanced statistical analyses. Count data were summarized as frequencies and continuous variables as means [H11006] standard deviations. A paired t test was used to evaluate changes in continuous variables. One-year primary patency, primary-assisted patency, secondary patency, and limb salvage were calculated by the Kaplan-Meier approach. All dependent variables were screened individually with the outcome, and those that met a conservative requirement (P < .30) were then moved forward to be included in a multivariate model. Multivariable Cox proportional hazards regression was used to develop predictive models. Patients presenting with ischemic rest pain were not included in the modeling for wound healing, but were included in the modeling for limb salvage. Statistical Analysis Software (SAS) version 9.2 (Cary, NC) was used for the statistical calculations.

RESULTS

A total of 123 limbs in 111 patients underwent tibial artery endovascular interventions. The mean age was 74 ±11.2 years, and 62% of patients were males. Comorbidities and risk factors are listed in Table I.

Table I.

Patient demographics

Characteristic	Frequency (n) or mean ± SD (median, range)
Mean age (years)	74.1 ± 11.2 (49-95)
Male	61%(68)
Diabetes mellitus	67% (74)
Chronic renal insufficiency	55% (60)
End-stage renal disease/dialysis	21% (24)
Hypertension	93% (102)
Hyperlipidemia	68% (72)
Prior coronary artery bypass graft	36% (39)
Coronary artery disease	72% (77)
Congestive heart failure	42% (43)
Angina	8% (8)
Unstable angina	1% (1)
History of myocardial infarction	38% (36)
Chronic obstructive pulmonary disease∥	19% (19)
Cancer∥	13% (13)
History of tobacco use†	51% (52)
Smoking status†
Never	49% (50)
Former	43% (44)
Current	9% (9)

^∥n = 101

^†n = 103

All interventions included a tibial artery angioplasty with or without proximal intervention. All patients were treated for critical limb ischemia; 102 limbs (83%) were treated for tissue loss (Rutherford class 5 and 6 disease). Fifty limbs (41%) underwent isolated tibial procedures, while 73 limbs (59%) had concurrent ipsilateral superficial femoral and/or popliteal artery intervention. Intervention on >1 tibial vessel was performed on 20% of limbs, and 14% included selective tibial laser atherectomy. The types and levels of primary interventions performed are outlined in Table II. Moreover, 62% (76) of tibial interventions were done for stenoses and 38% (47) for occlusions. In vessels treated for total occlusion, 68% were categorized as short (less than 1/3 the length of the vessel), 17% were categorized as medium (1/3 to 2/3 the length of the treated vessel), and 14% were categorized as long (>2/3 the length of the treated vessel).

Table II.

Summary ofprimary interventions

Primary interventions
Levels treated
Superficial femoral artery popliteal and tibial	27% (33)
Superficial femoral artery and tibial	14% (17)
Popliteal and tibial	19% (23)
Tibial only	41% (50)
Interventions	Percutaneous transluminal angioplasty	Stent	Laser atherectomy
Superficial femoral artery	100% (50)	68% (34)	6% (3)
Popliteal artery	100% (56)	25% (14)	16%(9)
Tibial vessels	100% (123)	4% (5)	14% (17)

Prior to intervention, 44% of the limbs treated had a poor popliteal and tibial runoff score, 56% were compromised, and none of the limbs were in the good range. Postprocedurally, this improved to no limbs in the poor range, 88% in the compromised range, and 12% in the good range (Table III). The mean distal popliteal and tibial runoff score improved from 11.8 ± 3.6 preprocedure to 6.7 ± 1.6 postprocedure (P < .001), and this improvement was similar for patients presenting with either rest pain or tissue loss. The mean ABI increased from 0.61 ± 0.26 preintervention to 0.85 ± 0.22 postintervention (P < .001).

Table III.

Distal popliteal and tibial runoff scores for limbs undergoing tibial artery interventions

	Category
Popliteal-tibial runoffscore	Good	Compromised	Poor	Mean
Preprocedure	0	56%	44%	11.8 ± 3.6
Postprocedure	12%	88%	0	6.7 ± 1.6

Periprocedural complications included groin hematoma (2.5%), pseudoaneurysm formation (0.8%), acute renal failure (0.8%), and the 30-day mortality was 1.7% (Table IV). No patient developed any procedure-related embolic complications. Two patients died within 30 days of their procedure, one from complications of a groin hematoma after an antegrade access that required operative repair and resulted in multisystem organ failure. The other death was secondary to in-dwelling line infection and overwhelming sepsis. The overall survival for the cohort was 83.7% at a mean follow-up time of 6.8 months.

Table IV.

Complications and length of stay for patients undergoing tibial artery interventions

Complication	% (n)
Hematoma	2.5% (3)
Pseudoaneurysm	0.8% (1)
Wound infection	0% (0)
Thrombosis	0% (0)
Acute renal failure	0.8% (1)
New or acute heart disease	0% (0)
Death
Within 30 days	1.7% (2)
Procedure-related	0.8% (1)
Mean length ofstay (days)	3.1 ± 5.4 (0-40)
Intensive care unit	4.1% (5)

At 1 year, the primary patency rate was 33%, and the assisted primary patency rate, as maintained by additional endovascular procedures, was 50%. Similarly, secondary patency rates as maintained by additional endovascular interventions were 56% (Fig 1).

Fig 1.

Patency in limbs undergoing tibial interventions (n = 121).

During the course of follow up, 39% of the patients treated for tissue loss had improvement in their wounds (mean follow up, 4.4 ± 4.8 months), while 41% had complete healing with a mean time to healing of 10.7 ± 7.4 months. TAEI in an isolated peroneal artery (odds ratio [OR] = 7.80; P = .01) as well as the need for hemodialysis (hazard ratio [HR] = 5.63; P = .04) were associated with impaired wound healing, whereas multi-level intervention (HR = 2.1; P = .009) and selective tibial laser atherectomy (HR = 3.1; P = .01) were predictors of wound healing. Overall, the limb salvage rate at 1 year was 75% (Fig 2). Factors associated with limb loss at 1 year included chronic renal insufficiency (HR = 5.73; P = .03) and pedal intervention (HR = 13.75; P = .04). The limb salvage rate at last follow up for those limbs that underwent an isolated peroneal intervention versus other tibial intervention was 74% versus 80% (P = .16). The limb salvage in patients with end stage renal disease was 54% at the last follow up, and was 82% for nondialysis–dependent patients (P = .15). Diabetes, smoking, statin therapy, and revascularization of >1 tibial vessel had no impact on limb salvage or wound healing. Patients who presented with ischemic rest pain were not included in the analysis of factors affecting wound healing; however, they were included in the analysis of limb loss.

Fig 2.

Cumulative limb salvage in at-risk limbs.

Life table analysis revealed a reintervention rate of 50% at 1 year. Thirty-three patients underwent reintervention during their follow-up; the main indications were failure of wounds to heal (58%) and duplex evidence of re-stenosis in the previously treated segment (24%; Table V). The types of repeat interventions performed to maintain primary-assisted and secondary patency are summarized in Table VI. Additionally, seven patients required subsequent bypass, one of which went on to major amputation. These patients are summarized in Table VII. Three other patients who had failed tibial re-intervention with no target for a bypass progressed to major amputation. The mean preprocedure popliteal tibial runoff score for the four patients who progressed to major amputation (including the one patient who had a subsequent bypass) was 14.6, and this improved to 8.9 after their initial intervention. Of the patients who failed re-intervention before progressing to amputation, none of them lost their target as a result of the intervention. These patients were not candidates for a bypass due to the lack of a target vessel and were offered an anatomically high-risk endovascular intervention for limb salvage as a last resort effort.

Table V.

Indications for reintervention on limbs undergoing tibial artery interventions

Indication	N = 33
Failure of wound healing	58% (19)
Duplex evidence of recurrent stenosis	24% (8)
Recurrent ulceration	6% (2)
Recurrent rest pain	6% (2)
Other	6% (2)

are summarized in Table VII. Three other patients who had failed tibial re-intervention with no target for a bypass progressed to major amputation. The mean preprocedure popliteal tibial runoff score for the four patients who pro-

Table VI.

Repeat interventions for maintenance of patency (36 interventions in 33 limbs)

	Primary assisted patency (n = 25) % (n)	Secondary patency (n = 11) % (n)
Level of Intervention
Superficial femoral artery	40% (10)	36% (4)
Popliteal artery	44% (11)	46% (5)
Original tibial vessel	88% (22)	91% (10)
Alternate tibial vessel	32% (8)	18% (2)
Type of Intervention
Angioplasty	96% (24)	100% (11)
Laser atherectomy	16% (4)	27% (3)
Stenting	28% (7)	46% (5)
Cryoplasty	12% (3)	0

Table VII.

Summary of patients converted to surgical bypass

Patient	Initial intervention	Bypass	Outcome
1	PTA PT and plantar	Pop-AT bypass	Healed toe amputation
2	PTAPT	Pop-DP bypass	Healed toe amputation
3	PTA peroneal	Pop-Plantar bypass	Healed ulcer
4	PTA/stent SFA and laser/PTA TPT and peroneal	Fem-AT bypass	Above knee amputation
5	PTA AT	Pop-DP bypass	Healed ulcer
6	PTA pop/TPT/peroneal	Fem-Peroneal bypass	Healed toe amputation
7	Laser/PTA peroneal	Pop-Peroneal bypass	Healed TMA

AT, Anterior tibial artery; DP, dorsalis pedis artery; Fem, femoral; Pop, popliteal; PTA, percutaneous transluminal angioplasty; TMA, transmetatarsal amputation; TPT, tibial-peroneal trunk.

DISCUSSION

Endovascular interventions for the treatment of critical limb ischemia have become the first-line approach in many centers.^2,10,11 Tibial artery endovascular interventions in the setting of CLI have been extensively described with mixed results.^1,12 A recent meta analysis by Romiti et al of infrapopliteal angioplasty for the treatment of CLI showed 1-year primary and secondary patency rates of 58% and 68%, respectively, with a limb salvage rate of 86% and patient survival of 98%.¹³ In the current study, we report a primary patency at 1 year of 33% with a secondary patency of 56% and a limb salvage rate of 75%. However, the majority of the limbs (83%) were treated for tissue loss with poor runoff scores and with limited target vessel options for salvage with conversion to bypass, which may in part explain the somewhat lower rate of limb salvage in the current series. Additionally, the lower rates of limb salvage are likely due in part to the fact that several patients were offered an endovascular intervention as a last resort effort, and most of them did not have adequate pedal runoff.

In patients treated for tissue loss, complete wound healing was achieved in 41% of cases with a mean time to healing of 10.7 months. Additionally, 39% had improvement in their wounds at last follow up, but 19% had worsened tissue loss or required major amputation, while 1% remained unchanged. These rates of wound healing and time to healing with TAEI are similar to those reported by other groups.⁵ Giles et al recently reported on their experience with infrapopliteal angioplasty for CLI in 176 limbs. At a mean follow up of 12 months, they reported complete healing or improvement in 57% of limbs, with stable wounds in 22% and worsening of the wound in 21%.⁵

One factor that may affect the rate of wound healing is the degree of tissue loss present, in that a large volume of tissue loss may predict failure of TAEI to achieve wound healing. However, specific wound size information was not available on enough patients to analyze this factor in the current study. An additional confounder that may explain the low healing rates is the length of follow up. In the healed group, the mean time to healing was 10.7 months, while the group that showed improvement in the wounds had a mean follow up of 4.4 months. It is the authors’ belief that with longer follow up, close surveillance, and repeat interventions as indicated, improved wound healing rates may be attainable.

In this series, multi-level interventions favorably affected wound healing when compared with isolated TAEI, but did not have a significant effect on overall limb salvage. This may reflect to some degree an underestimation and undertreatment of proximal disease in the isolated tibial intervention group, where aggressive treatment of proximal disease in the multi-level group would provide better flow to the threatened foot. Although wound healing was not directly evaluated in the meta-analysis by Romiti et al, there was a trend toward better primary patency and limb salvage with multi-level interventions compared with crural angioplasty alone.¹³ Wound healing was not included in the report by Sadek et al, but they reported trends toward improved limb salvage and primary patency, as well as significantly improved secondary patency with multi-level interventions involving the tibial vessels.¹⁰ One explanation offered was that patients with single-level disease may exhibit locally increased atherosclerotic burden compared with patients with multilevel disease, and that this may result in increased primary failure and need for secondary interventions in patients with single-level infrapopliteal disease.

Selective laser atherectomy was found to have a positive impact on wound healing, although it resulted in similar patency rates. This technology has been utilized with mixed early results with limb salvage rates at 12 months as low as 55% in some reports.¹⁴ In the current study, the use of laser atherectomy was at the discretion of the operating surgeon, and was primarily used for the debulking of calcified tibial lesions prior to angioplasty. This was most commonly done for orifice lesions, bifurcation lesions, and as an assist to cross total occlusions. Importantly, laser atherectomy was not used at any level as a stand-alone intervention, and was followed by balloon angioplasty of the treated segment. The number of limbs in the current series treated with atherectomy is small, and the observed positive impact on wound healing may be a result of selection bias of the treating surgeon as it was utilized in specific anatomic situations, and it was used primarily to debulk calcified or orifice lesions prior to balloon angioplasty. Given this bias, the beneficial effects observed may not be widely applicable and will need to be validated with a larger sample size.

Only patients who had completed successful endovascular interventions were included in the study. Overall, 65% of the treated limbs were felt to have angiographically established in-line flow to the foot, and the remaining 35% with an isolated peroneal artery intervention had either a discontinuous or no pedal runoff, or communication to a pedal runoff via peroneal collaterals at the ankle. The distribution of in-line pedal flow was not different between the groups that attained complete wound healing, had partial healing, had worsening wounds, or that progressed to major amputation. Among the factors found to negatively affect wound healing included TAEI on an isolated pero-neal artery. This may be related to the fact that, even with successful treatment of the peroneal artery, in-line flow to the foot is not established. Although it has been shown that surgical bypass to the peroneal artery is an acceptable option with good wound healing and limb salvage rates,^15,16 and that there is evidence that bypass to the peroneal artery is not hemodynamically inferior to other tibial artery bypass grafts,¹⁷ it remains to be determined if the same is true for endovascular interventions. Dosluoglu et al reported on their experience with endovascular interventions for limb salvage, comparing patients with isolated peroneal artery runoff and those with other tibial runoff. They noted similar patency and limb salvage rates between the groups.¹⁸ They concluded that endovascular revascularization of an isolated peroneal runoff resulted in acceptable patency and limb salvage rates in patients presenting with tissue loss. However, it is important to note that less than half of the patients in their peroneal runoff group had isolated infrapopliteal interventions, which might have contributed to their positive outcomes.¹⁸ In contrast all patients in our study underwent isolated peroneal interventions, and interventions on an isolated peroneal runoff were done when there were no other options to improve outflow to the foot. As such, the association of an isolated peroneal intervention with poor wound healing is likely explained by suboptimal pedal runoff, and possibly a greater local atherosclerotic burden. In support of these findings, the updated TASC-II guidelines state that for the endovascular treatment of infra-popliteal disease, angioplasty may be indicated for limb salvage, and the treatment of tibial artery occlusion should be reserved for cases in which in-line flow into the pedal vasculature can be established.⁴

However, peroneal interventions in our series did not seem to affect limb salvage, and it may be that interventions on an isolated peroneal runoff can result in limb salvage but require longer follow up and prolonged wound care, since pedal perfusion is likely improved via collateral pathways at the ankle.

In addition to renal insufficiency, interventions at the pedal level were associated with an increased risk of limb loss. On the other hand, patients on dialysis had similar limb salvage as nondialysis patients. However, the number of dialysis patients treated in this series was relatively small.

There is very little in the literature about the outcomes of isolated pedal intervention and their effect on wound healing and limb salvage. Although the numbers of such patients treated in this series is small, it was predictive of limb loss, and did not have a favorable effect on wound healing. The association of pedal intervention with poor outcomes is likely explained by poor patency and diffuse inframalleolar small vessel disease, and may be a reflection of a combination of overall greater burden of disease and poor pedal runoff.

Alternatively, pedal bypass for limb salvage has been shown to be an effective intervention for limb salvage, with a recent meta-analysis showing 1-year and 5-year limb salvage rates of 88% and 75%, respectively.¹⁹ Despite the small numbers, and given our findings of impaired limb salvage with pedal angioplasty, it is our current belief that pedal bypass should be done preferentially over endovascular interventions for anatomically and physiologically suitable patients who require revascularization for isolated distal tibial or pedal disease. Nevertheless, it is still our practice to perform pedal interventions for patients who are at a physiologically high risk for bypass.

Similarly, although our current report describes the limitation of interventions on isolated peroneal runoff vessels for wound healing, we do offer these interventions for patients presenting with tissue loss who have good flow into the foot via collateral pathways at the ankle. Patients with a patent pedal vessel and poor collateralization from the peroneal artery are preferentially offered a pedal bypass if a vein conduit is available, and the patient has no prohibitive co-morbidities.

Although we are aggressive in our attempts at limb salvage, in patients with extensive tissue loss and nonsalvageable limbs, such as dialysis patients with extensive heel gangrene, a primary amputation is considered. Likewise, nonambulatory patients with knee contractures and extensive tissue loss are also offered a primary amputation.

In an attempt to quantify the disease burden treated at the popliteal and tibial levels, we looked at preprocedure angiograms and operative notes to assign a distal popliteal and tibial runoff score. This was done according to a modification of the SVS criteria such that a higher score implies worse runoff and limb salvage rates.⁹ In the current series, the mean distal popliteal and tibial runoff score improved postintervention. This can be explained by the fact that popliteal disease, when present, is weighted more significantly than individual tibial vessels, and that the majority of the tibial interventions were done on a single tibial vessel in an attempt to obtain in-line flow to the foot. However, neither the preprocedural, postprocedural, or change in runoff score were found to be associated with wound healing or limb salvage rates. Moreover, 62% of tibial interventions were done for stenoses and 38% for occlusions, and there were no differences found in wound healing or limb salvage outcomes between these two groups. Intuitively, although it would be expected that total occlusions, particularly longer occlusions, would predict worse outcomes, the similar outcomes observed may be due to the small number of patients and lack of power to detect differences in subgroup analyses.

In the TASC II document, a classification for tibial disease was not included, despite the fact that it was present in the initial TASC document. The updated TASC-II guidelines state that for the endovascular treatment of infrapopliteal disease, angioplasty may be indicated for limb salvage, but the classification was not updated given the lack of supporting literature.⁴ For that reason, the current study did not analyze data by TASC classification and used the tibial runoff score instead.

All tibial interventions were performed with the intent of establishing in-line flow into the foot. Several patients underwent interventions on multiple tibials in an attempt to maximize perfusion to the foot at the discretion of the operating surgeon. Intervening on multiple tibial vessels was not found to have any benefit in terms of wound healing or limb salvage in the current analysis. The effect of intervening in the same tibial bed at multiple levels was not analyzed and cannot be correlated with limb outcomes.

Although diabetics constituted 67% of our patients, this did not affect limb salvage or wound healing. There have been mixed data reported in the literature on the impact of diabetes on the outcomes of endovascular treatment of CLI. Diabetes has been associated with impaired primary patency rates requiring higher rates of reintervention in the treatment of infrainguinal occlusive disease in patients presenting with both claudication and CLI.² Similarly, other series reported lower rates of limb salvage in diabetics treated with endovascular interventions for CLI despite attaining equivalent patency rates.²⁰ However, others showed that, although diabetics display decreased primary patency, with appropriate reintervention, improved secondary patency and limb salvage rates can be attained.²¹ In the current study, diabetes was not a negative predictor of wound healing or limb salvage. This likely reflects a combination of our aggressive surveillance and reintervention protocols and possibly a patient selection bias as some patients were referred by other surgeons only when they were deemed to be at high risk for bypass, while other were preferentially treated with a bypass.

There are several limitations to this study, primarily inherent to its retrospective design. Additionally, as there currently are no consensus guidelines to classify tibial interventions, the lack of reporting standards may yield mixed results from different studies. In this retrospective single institutional review, the approach to treatment of CLI was left to the discretion of the operating surgeon. Where an endovascular first approach was adopted by the majority, others utilized TAEI only in patients with favorable anatomy or those deemed to be at high risk for surgical bypass. Additionally different revascularization modalities and techniques were operator-dependent, therefore resulting in a heterogenous patient population with different comorbidities and disease distribution.

Finally, longer follow up is needed to determine the durability of TAEI and their long term effect on wound healing and limb salvage, and larger cohort of patients are required to compare the efficacy of the different available endovascular modalities.

CONCLUSIONS

Tibial artery endovascular interventions result in acceptable rates of limb salvage and wound healing in an often challenging patient population with compromised tibial outflow. In addition, these interventions can be carried out in situations when surgical bypass is not possible due to severe medical risk factors. Pedal endovascular interventions had a negative impact on limb salvage, and patients with pedal disease distribution may be better treated with a surgical bypass if medically and anatomically suitable. Revascularization of multiple tibial vessels showed no additional benefit over single tibial interventions, and isolated peroneal interventions were found to be associated with impaired wound healing. However, these findings will need to be validated in larger series.