Surgical bypass grafting for critical limb ischemia — target selection, conduit choice, anastomosis, surveillance, and long-term patency
Below-knee bypass is offered when endovascular angioplasty or stenting has failed, is technically unsuitable, or is unlikely to durably relieve tissue-threatening ischemia. Selecting the correct distal target — the tibial or pedal artery to which the graft will be sewn — is the single most consequential decision in the operative plan, because runoff quality below the anastomosis is the strongest predictor of whether the reconstructed limb will heal.
Peripheral arterial disease (PAD) becomes critical limb ischemia (CLI, now often termed chronic limb-threatening ischemia, CLTI) when rest pain, non-healing ulceration, or gangrene develops from inadequate arterial perfusion. The Global Vascular Guidelines and BEST-CLI trial data support an "endovascular-first" approach for many patients, but surgical bypass remains preferred or necessary when:
• Long-segment (>25 cm) infrapopliteal occlusive disease unsuitable for angioplasty • Prior failed endovascular intervention with recurrent occlusion • Heavily calcified, non-crossable, or non-dilatable tibial segments • A good-quality single autologous saphenous vein conduit is available in a patient fit for open surgery • BEST-CLI (NEJM 2023) showed that among patients with a usable single-segment great saphenous vein, surgical bypass produced superior major-adverse-limb-event-free survival compared to endovascular therapy
The operative plan begins with a full below-knee angiogram or CT angiogram mapping the tibial trifurcation and pedal arch, since the distal target and expected outcome depend entirely on what vessel bed remains patent.
The popliteal artery trifurcates into the anterior tibial artery, the tibioperoneal trunk (which divides into the posterior tibial and peroneal arteries). Each has distinct anatomic and functional implications for bypass targeting:
Posterior tibial artery — travels posteromedially behind the medial malleolus to become the plantar arteries; it is usually the most favorable target because it directly perfuses the plantar forefoot, the most common site of CLI ulceration, and is often more resistant to diffuse calcific disease than the anterior tibial.
Anterior tibial artery — travels through the interosseous membrane to become the dorsalis pedis; a good target for dorsal foot wounds and is often the last vessel to occlude in diffuse three-vessel disease, making it useful when it is the only patent option.
Peroneal artery — runs deep in the posterior compartment and does not directly reach the foot, relying on collateral communication (via the anterior and posterior communicating branches) to perfuse the pedal arch; it is chosen when both tibial vessels are occluded, and outcomes depend heavily on the quality of these collaterals.
The angiosome model divides the foot and leg into three-dimensional blocks of tissue fed by a specific source artery. Direct revascularization — restoring flow through the artery that anatomically supplies the wound bed — is associated with faster wound healing and lower reintervention rates than indirect revascularization through collaterals alone, though when a direct target is unavailable, a good-quality indirect target with robust collaterals remains an acceptable choice.
Runoff is typically graded using an angiographic scoring system (e.g., Society for Vascular Surgery runoff score) that accounts for the number of patent tibial vessels reaching the foot and the continuity of the pedal arch. A patient with 3/3 patent tibial vessels and an intact pedal arch has dramatically better expected graft patency than one with a single, diseased outflow vessel — this is why runoff score is one of the two primary sliders driving the patency model in this simulation.
The choice of bypass conduit is the second major determinant of long-term success, and the evidence is unambiguous: for below-knee bypass, autologous great saphenous vein produces meaningfully better patency than any prosthetic alternative. Conduit selection is nonetheless a balance between biology and availability — not every patient has usable vein, and every graft, vein or prosthetic, still depends on adequate inflow and outflow to remain open.
Autologous vein is biologically compatible living tissue: its endothelial lining resists thrombosis, it can slowly remodel and dilate in response to flow, and it is far less prone to the pseudo-intimal hyperplasia that plagues synthetic grafts at the anastomotic suture line. Below the knee, where target vessels are small-caliber (often 2–3 mm) and flow rates are lower than above the knee, these advantages become decisive: prosthetic grafts thrombose at rates several-fold higher than vein at these small distal targets.
The great saphenous vein (GSV) is the workhorse conduit, harvested either in reversed configuration (proximal vein sewn distally so valves do not obstruct flow), non-reversed with valve lysis (in situ technique, preserving a more size-matched taper), or translocated. Vein diameter of at least 3 mm along its length, absence of extensive varicosities or sclerosis, and continuity without excessive branching are the key criteria for a "good-quality" conduit.
Approximately 20–30% of patients presenting for below-knee bypass lack adequate ipsilateral great saphenous vein — from prior harvest (e.g., coronary bypass), vein stripping, chronic venous disease, or insufficient caliber. In these cases, options in order of preference include:
1. Alternative autologous vein: contralateral GSV, lesser saphenous vein, or arm vein (cephalic/basilic), sometimes spliced together as a composite conduit 2. Cryopreserved allograft vein: intermediate patency, used when no autologous source exists and prosthetic is relatively contraindicated (e.g., infected field) 3. Expanded polytetrafluoroethylene (ePTFE) or Dacron prosthetic graft: most common fallback; patency can be improved with an adjunctive vein cuff or patch (Miller cuff, Taylor patch) at the distal anastomosis, which reduces compliance mismatch and intimal hyperplasia
Prosthetic grafts remain a reasonable option in patients with limited life expectancy, no usable vein, or urgent need for revascularization where harvesting vein would add prohibitive operative time.
Landmark comparative series consistently show roughly a 25–30 percentage-point gap in 5-year primary patency between vein and prosthetic conduit for infrapopliteal bypass — vein around 60–70%, prosthetic around 30–40% — a difference large enough that vein is considered the standard of care whenever available.
Once inflow source, distal target, and conduit are selected, the operation proceeds to physical construction of the bypass: tunneling the conduit along an anatomic or subcutaneous path, and hand-sewing the proximal and distal anastomoses that will carry pulsatile arterial flow around the diseased or occluded native segment.
Inflow is chosen as the most distal healthy arterial segment to shorten conduit length and preserve future bypass options: the above-knee or below-knee popliteal artery is favored when patent, since using the common femoral artery as inflow when the superficial femoral/popliteal segment is normal simply lengthens the graft unnecessarily and wastes vein.
The conduit is tunneled either anatomically (following the native vessel course, deep to the sartorius and gastrocnemius) or subcutaneously, depending on surgeon preference, prior scarring, and skin quality overlying the leg. A tunneling device creates a smooth subfascial or subcutaneous track free of kinks or compression, since any external compression along the tunnel predisposes to early graft thrombosis.
Both anastomoses are constructed end-to-side using fine monofilament suture (typically 6-0 or 7-0 polypropylene) in a continuous running technique, under loupe or microscope magnification for the small-caliber distal anastomosis:
Proximal anastomosis: sewn to the inflow artery (e.g., popliteal), typically at a gentle oblique angle, with the heel and toe of the anastomosis carefully constructed to avoid narrowing.
Distal anastomosis: the technically demanding step, given target vessel diameters often only 2–3 mm. An elliptical arteriotomy is fashioned and the graft spatulated to create a wide, low-angle (30–45°) anastomosis that minimizes turbulence and the compliance mismatch known to drive intimal hyperplasia at the suture line. Adjuncts such as a vein (Miller) cuff or patch are used when a prosthetic conduit is anastomosed to a small tibial vessel, interposing a short vein segment to soften the transition in compliance between the stiff synthetic graft and the native artery.
Completion assessment: intraoperative duplex ultrasound or angiography confirms a widely patent anastomosis, absence of a technical flow-limiting defect, and triphasic or biphasic Doppler signal in the pedal vessels before closure.
Once both anastomoses are complete and clamps released, pulsatile flow is restored through the graft to the foot for the first time. A palpable pulse or strong Doppler signal at the distal anastomosis and in the pedal vessels, along with a visibly perfused, warm foot, are the immediate intraoperative markers of technical success. Ankle-brachial index (ABI) is typically rechecked postoperatively and should show substantial improvement compared to the preoperative value, though a below-knee bypass to a diseased distal bed may not fully normalize ABI even when technically successful — clinical perfusion (capillary refill, wound bed appearance, toe pressures) is followed alongside the numeric index.
A structurally successful bypass graft is not a "fire-and-forget" repair. Intimal hyperplasia — a fibrotic overgrowth of smooth-muscle and matrix at the anastomoses or within a vein conduit — narrows the graft lumen over the first months after surgery in a meaningful subset of patients. Because focal stenosis is treatable while a fully thrombosed graft is far harder to salvage, structured duplex ultrasound surveillance is standard of care after infrainguinal bypass.
The dominant mode of vein graft failure in the first 1–2 years is not atherosclerosis but intimal hyperplasia — an overzealous smooth-muscle-cell and extracellular-matrix healing response at sites of vein wall injury (valve sites, anastomoses, or areas of endothelial trauma during harvest). Left undetected, a focal hyperplastic stenosis silently narrows the graft until it thromboses, typically presenting as an acute, symptomatic occlusion.
Duplex ultrasound surveillance detects these lesions while the graft is still patent, by measuring peak systolic velocity (PSV) at intervals along the entire length of the conduit and calculating the velocity ratio (PSV at the stenosis divided by PSV in a normal adjacent segment). A rising PSV or ratio over serial studies — even before a fixed threshold is crossed — flags a graft "at risk" and prompts closer monitoring or intervention.
A typical surveillance schedule begins with a baseline duplex within the first month, then every 3 months through the first year, and every 6 months thereafter — the interval front-loaded to the period of highest hyperplasia risk. Each study records graft diameter, PSV at multiple segments, and ankle-brachial index.
Commonly used revision thresholds: • PSV within the graft <45 cm/s, or a resting ABI drop of >0.15 from postoperative baseline, suggests global low flow and graft-at-risk physiology • Focal PSV >300 cm/s with a velocity ratio >3.5 at a specific segment indicates a hemodynamically significant focal stenosis • Progressive narrowing on serial studies, even below threshold, warrants closer follow-up or elective angiographic evaluation
When a significant lesion is found, revision — via balloon angioplasty of the focal stenosis, or open patch/interposition revision for vein grafts — is performed electively, restoring durable flow before thrombosis occurs. This surveillance-and-revision strategy has been shown to significantly improve secondary (assisted) patency and limb salvage rates compared to a "watch and wait for symptoms" approach.
A stenotic but still-patent graft treated electively has a very high (>80%) chance of returning to durable long-term patency after revision. Once a graft has thrombosed, thrombectomy or thrombolysis succeeds in only a minority of cases, and many occluded grafts require an entirely new bypass — this asymmetry is the entire rationale for surveillance.
Graft patency is reported using three overlapping definitions that together describe the durability of a bypass: primary patency (never occluded or reintervened upon), assisted-primary patency (kept open by elective revision of a stenosis before it ever thromboses), and secondary patency (reopened successfully even after a thrombotic episode). Comparing vein and prosthetic conduits across these three measures, over 1 and 5 years, is the clearest way to communicate expected durability to a patient.
Vascular literature standardizes outcome reporting into three tiers, each answering a slightly different clinical question:
Primary patency: the graft has remained continuously open, with no occlusion and no intervention of any kind, from the day of surgery. This is the strictest and most conservative measure.
Assisted-primary (primary-assisted) patency: the graft has never occluded, but has required one or more elective procedures (angioplasty, patch revision) to correct a developing stenosis found on surveillance — exactly the scenario surveillance is designed to enable.
Secondary patency: the graft did occlude at some point, but was successfully reopened (via thrombectomy, thrombolysis, or a revision/jump graft) and remains functional. Once a graft has required this level of intervention its long-term durability is generally lower than one that only needed assisted-primary care.
All three numbers are typically reported at 1-year and 5-year time points, and all three are consistently and substantially better for vein than for prosthetic conduit at infrapopliteal targets.
Beyond conduit type, several modifiable and non-modifiable factors independently predict graft failure:
• Poor distal runoff: fewer than 2 patent tibial vessels or an incomplete pedal arch is one of the strongest predictors of both early thrombosis and late failure, regardless of conduit • Continued smoking: tobacco use after bypass surgery accelerates intimal hyperplasia and disease progression in both the graft and native vessels, and is consistently associated with markedly worse patency and higher amputation rates • Poor-quality or small-caliber vein (<3 mm), or a composite/spliced vein conduit, underperforms a single-segment, adequate-caliber GSV • Diabetes and end-stage renal disease/dialysis dependence are associated with more extensive tibial disease, calcification, and reduced patency and limb salvage rates • Redo bypass (after a prior failed graft) generally has lower patency than a first-time bypass, reflecting more advanced, harder-to-treat disease
Modifying what can be modified — smoking cessation, optimal antiplatelet/statin therapy, glycemic control, and adherence to the surveillance schedule — measurably improves the odds of long-term limb salvage.
For a patient with good-quality single-segment saphenous vein and reasonable runoff (2–3 patent tibial vessels), a below-knee bypass can be expected to achieve roughly 80–85% primary patency at 1 year and 60–70% at 5 years, with secondary patency (accounting for successful revisions) climbing higher still.
For a patient requiring prosthetic conduit to a single, marginal tibial target, expected 1-year primary patency drops to roughly 60–70%, and 5-year primary patency may fall to 30–40%, though adjuncts like a distal vein cuff and rigorous surveillance can narrow this gap somewhat.
These figures — reproduced by the interactive patency curves in this simulator — underscore why vascular surgeons pursue autologous vein whenever feasible, why runoff assessment drives target selection from the very first angiogram, and why surveillance remains a lifelong commitment for anyone who has undergone infrainguinal bypass.
Across the literature, the combination of best-case conduit (single-segment great saphenous vein) and best-case runoff (3 patent tibial vessels with an intact pedal arch) can approach 90% 1-year and 70%+ 5-year primary patency — while worst-case combinations (prosthetic conduit, single diseased runoff vessel, continued smoking) can fall below 50% at 1 year and below 20% at 5 years. The two sliders in this simulator span exactly that range.