Peripheral vascular intervention is better than tibial bypass for critical limb ischemia

Critical limb ischemia (CLI) carries a prognosis similar to many cancers because of the malignant nature of atherosclerotic disease. The majority of CLI patients have multilevel disease, both above and below the knee. Approximately one-third have isolated occlusive disease in the below-knee (BK) arteries. The prognosis for those with isolated BK disease is worse than those with multilevel disease. ¹ Patients with isolated BK disease tend to be very old, very thin, very sedentary, and/or have renal insufficiency.^1–7

With that background, the vascular surgery group from Johns Hopkins University has reached into the Society for Vascular Surgery Vascular Quality Initiative (SVS-VQI) registry to review the outcomes of patients treated for CLI with revascularization of BK disease. ² They reviewed procedures entered into the registry from 2008 to 2014 that included either lower extremity bypass surgery (LEB, n=500) or peripheral vascular intervention (PVI, n=2066). They excluded any patient who required more proximal artery reconstruction and used data only from institutions that achieved at least 50% follow-up during the 9–15-month post-treatment window. The main finding was that PVI had a higher patency compared to LEB (81% vs 73%; p<0. 001) at 1 year. Harder endpoints of major amputation (12% vs 14%; p=0.18) and mortality (6% vs 4%; p=0.15) showed no significant difference based on revascularization strategy.

The study subjects represent a high-risk population with unfavorable traditional risk factors that included diabetes (71%), tissue loss (77%), and dialysis (18%). The high amputation and mortality rates seen are in keeping with this subset of CLI patients (see Table 1). Patency of revascularization was determined with arteriography (10.6%), duplex ultrasound (35.3%), palpable pulse or graft (26.5%), or an increase in ABI of >0.15 (7.5%). The main conclusion was that PVI ‘can be considered as first-line approach in patients presenting with diabetes and/or tissue loss’. The strength of this conclusion is offset by the inconsistent objective analysis for patency coupled with less than optimal follow-up. The authors did analyze separately only those with duplex ultrasound follow-up. They showed a significant difference in the primary patency of PVI over LEB (73% vs 61%, p<0.001); however, the reported patencies were 8–12% less when using duplex ultrasound compared to less objective measures.

Table 1.

Summary of data from recently published studies generated from Society for Vascular Surgery Vascular Quality Initiative (SVS-VQI) for patients treated with critical limb ischemia.

SVS-VQI Study	Study range	Total no. of CLI pts studied	Bypass no.	% Of pts excluded due to lack of f/u	Procedure	Loss of primary patency or MALE	AFS rate	Death
			PVI no.
Hicks 2	2008–2014	2566	n=500 Bypass	n=5851 were excluded (~70%)	Tibial bypass	27% LPP	82%	4%
			n=2066 PVI		Tibial PVI	19% LPP	82%	6%
Brothers 5	2003–2014	4354	n=3036Bypass	42%	Tibial bypass	36% LPP;31% MALE	63%	23%
			n=1319PVI	75%	Tibial PVI	31% LPP;30% MALE	68%	14%
Siracuse 6	2010–2013	7897	n=3059Bypass	45%	SFA (~50%) + tibial (~50%)	~35% MALE	~89%	~5%
			n=4838PVI	56%	SFA + tibial	~35% MALE	~85%	~10%
Dermondy 7	2003–2013	3375	n=3375Bypass	NA	<50 years old	38% MALE	69%	6%
				NA	50–79 years old	30% MALE	75%	10%
				NA	>80 years old	31% MALE	59%	26%

All endpoints were determined at 1 year of follow-up.

CLI, critical limb ischemia; pts, patients; PVI, peripheral vascular intervention; f/u, follow-up; AFS, amputation free survival; LPP, loss of primary patency; MALE, major adverse limb events; SFA, superficial femoral artery; NA, not available/reported.

Registry data suffer from reporting biases, evolving treatments over the study time interval, and inconsistent data reliability. The assumption that these biases and inconsistencies affect subpopulations equally may not be true. Even data from audited randomized controlled clinical trials may be scrutinized for bias and inconsistency. The BASIL (bypass versus angioplasty in severe ischemia of the leg) trial provides the only randomized CLI data. ³ They compared two revascularization methods for CLI (open surgery and balloon angioplasty) and included 452 patients who were followed for 5.5 years. However, the BASIL trial only represents a minority of CLI patients because only 11% of CLI patients were randomized at the six leading centers, 70% had single-level percutaneous transluminal angioplasty (PTA) of the above-knee superficial femoral artery (SFA), with multilevel angioplasty in only 30%. In addition, only balloon angioplasty was used, no other PVI modalities. Bypass patients had higher patency for those who lived beyond 2 years (mortality rate 27% at 2 years); however, mortality in the surgical arm was higher than the PTA-treated patients at 7 months, and outcomes were only improved long term if the saphenous vein was used. The PTA-treated patients did better than those treated with synthetic bypass.

Registry data represent a broader patient base than can be studied in a randomized controlled trial that has strict inclusion/exclusion criteria. A registry can compare all patients with similar anatomy, not just a minority. Also, the revascularization techniques utilized in registries are more ‘real world’, and can include technologies such as atherectomy, bare metal stents, and drug-eluting stents. Finally, registry results are a reflection of non-randomized physician decision-making that hopefully centers on what is best for the patient when it comes to revascularization method.

Several other general observations and comments about the SVS-VQI registry can be made based upon the report of Hicks and colleagues in this issue of Vascular Medicine as well as previous registry-based publications (Table 1).^2,5–7 Firstly, the utilization of PVI outpaces LEB at least 4 to 1. No longer are bypasses the preferred treatment for CLI, even in the vascular surgery community. This dramatic shift has happened because endovascular procedures have become a regular and routine component of vascular surgery training. Secondly, PVI procedures are less time-consuming and easier on the patient as well as the practitioner compared to LEB.^5–7 They burn few bridges, even knowing that up to one-third of patients receiving LEB have previously had PVI. ⁷ Thirdly, Table 1 provides a summary of SVS-VQI publications that have reported 12-month outcomes of patients undergoing revascularization for CLI. While these publications differ in study dates, number of available patients with adequate follow-up, and types of procedures studied, they do provide some general impressions of the registry. In these SVS-VQI CLI reports, around 50% of the study subjects have follow-up within the 9–15-month window after revascularization, major adverse limb events occur in about one-third of patients, amputation-free survival is worse in older patients and independent of revascularization method, and mortality ranges from 4% to 26% at 1 year.^2,5–7

Why are these data important? Practitioners want to make the right decisions regarding revascularization for their patients (open bypass or catheter-based intervention); however, heterogeneity of the disease, impact of concomitant medical comorbidities, variability in the quality of revascularization (indirect versus direct; single vessel versus multi-vessel; balloon angioplasty versus stent; good vein conduit versus suboptimal vein conduit, etc.) challenge ANY study to provide patient-specific treatment dictums.

The best clinical decision-making is still made at the bedside, taking into account patient-specific variables, including: comorbidities, previous failed procedures, quality of the saphenous vein, and the patient’s wishes/preferences. Coupled with operator skill, experience, and integrity, good judgment should follow. This cannot be a ‘one size fits all’ mentality if the decision process is prudent.

In summary, recent publications from the SVS-VQI registry data should open some eyes to how treatment for CLI has changed. Both LEB and PVI are viable treatment options for this condition. The best choice for revascularization may depend more upon the ability to provide sustained functional recovery, independence of living, and quality of life, and less on predicting patency and mortality.