Obliterative atheromatous disease of the femoral and distal arteries does not exist in isolation but is part of a widespread vascular pathology; the cardiac, cerebral, and less commonly the mesenteric and renal circulations may be affected. Initial deposits of atheroma in the vessel wall are overlaid with hyaline collagenous material which projects into the arterial lumen. This plaque may ulcerate, leading to superimposed thrombosis which organizes and enlarges, further narrowing the vessel and causing turbulent flow. Turbulence accelerates the process leading to occlusion of the vessel.

The natural history of infrainguinal arterial disease is not a simple steady deterioration towards amputation; it is more often characterized either by stable intermittent claudication or even by symptomatic improvement as collateral channels enlarge. The risk of gangrene or pregangrene within a year of presentation is about 5 per cent, and about 2 per cent per annum thereafter. Of every 100 patients with claudication, approximately 40 will improve, 40 will remain unchanged, and 20 will require operation. However, the mortality rate of these patients is twice that of age and sex matched controls without peripheral vascular disease.

The three cardinal features of peripheral lower limb ischaemia are intermittent claudication, rest pain, and gangrene, representing an increasing degree of severity of ischaemia. Intermittent claudication is a cramp caused by inadequate oxygenation of muscle. It is initiated by walking and relieved by rest; generally the calf muscles are most affected. Claudication distance remains roughly the same unless the underlying condition deteriorates, although the symptom is more pronounced on hurrying or going uphill.

Rest pain occurs when the blood supply is so poor that tissue perfusion is inadequate even at rest. The pain classically affects the toes or forefoot (the most distal part of the limb) although in severe cases it may involve the whole foot or calf. It is usually first noticed in bed, when the patient is horizontal, the beneficial effect of gravity is removed and the foot is warmed, thereby increasing metabolism. External stimuli are reduced at night and so pain may be appreciated more readily. Rest pain is helped by hanging the leg out of bed, standing, or even walking. As ischaemia progresses the pain becomes continuous, requiring opiate analgesia for its control. The third clinical feature is gangrene. This is the end stage of ischaemia when the circulation is so poor that necrosis ensues. It usually begins distally in the foot (Fig. 1) 247.

In addition to the three cardinal features of ischaemia the concept of ‘critical limb ischaemia’ is useful in vascular surgery. This is defined as persistent rest pain requiring analgesia for more than 2 weeks, or ulceration, or gangrene of the foot, plus an ankle systolic pressure below 50 mmHg. In diabetics, owing to the unreliability of ankle pressure recording due to vascular calcification, the pressure criterion is replaced by absence of ankle pulses.

Assessment must establish the degree of ischaemia, whether it requires treatment and, if so, the most appropriate treatment. History and examination will usually identify the presence or absence of vascular disease and suggest its severity. Skin temperature, pallor on elevating the limb followed by dependent rubor, and the absence of pulses are particularly important features. The palpation of pulses should give the surgeon a rough idea of the site of arterial occlusion. More exact assessment requires a Doppler ultrasound probe and sphygmomanometer cuff to measure the highest opening systolic pressure of the three ankle arteries (Fig. 2) 248. This is divided by the higher of the brachial systolic pressures to obtain an ankle/brachial pressure index, the normal value of which is greater than 1. Values of 0.6 to 0.9 are typical of claudication, 0.3 to 0.6 of rest pain, and below 0.3 of incipient or actual gangrene. In some individuals with apparently normal values at rest, occult disease may be uncovered if the ankle/brachial index falls after exercise.

The need for intervention is easy to assess in those with mild claudication or critical ischaemia. Between these extremes a balance must be struck between the risk to life and limb from any proposed procedure and the compromised lifestyle which may result from conservative management. This balance will be affected by the patient's social circumstances and by the results of further investigations to assess fitness for operation. Although simple palpation of pulses usually allows the approximate level of occlusion or stenosis to be recognized, if surgical or radiological intervention is intended some form of imaging is required to provide more detail: the standard imaging technique is angiography (Fig. 3) 249. Radiographs are exposed after injection of radio-opaque contrast medium into the arterial tree through a fine catheter inserted via the femoral artery in the groin. Current techniques using non-ionic contrast media and narrow gauge catheters are safe, though invasive, and the angiogram remains the investigation of choice. Computerized (digital subtraction) angiography may be used as an adjunct to the basic conventional technique to highlight areas of special interest (Fig. 4) 250. This modern method may also be employed on its own in some cases, using either an intra-arterial or an intravenous injection. The advantage of the former is that very fine catheters may be used and small quantities of contrast media injected. The advantage of the latter is that the arterial tree does not have to be invaded at all. However the intravenous technique requires administration of a large volume of contrast agent, which debilitated patients may tolerate poorly. Furthermore, the image quality of intravenous digital subtraction angiography for the leg vessels is not particularly good.

It must be stressed that angiography should only be performed if intervention is intended. It allows an assessment of whether intervention is technically feasible and enables the most appropriate form of treatment to be chosen. The appearance of the aorta and iliac vessels is checked to confirm that there is no impairment of inflow to the leg. The sites of stenosis and occlusion in the leg arteries themselves are noted, and patency of the distal arteries (outflow) assessed. The usefulness of angiography in assessing patency of very distal arteries when an upstream occlusion is present has recently been questioned. The technique of pulse generated run-off may detect patent vessels at the ankle which have not been adequately demonstrated by angiography. A pneumatic cuff around the upper calf is mechanically inflated and deflated rapidly to generate a pseudopulse which is detected by a Doppler ultrasound probe over any patent ankle artery. In patients with critical ischaemia this may allow the possibility of a femorodistal bypass to be recognized in the presence of a negative angiogram.

Atheroma has a predilection for certain arterial sites and patterns of disease are recognizable: within the leg these common patterns are femoropopliteal disease and distal disease (Fig. 3) 249. Nevertheless, atheroma is never entirely limited to one site, although relieving an obstruction at one site may result in a general relief of symptoms. The most common site of stenosis or occlusion in the femoropopliteal segment is the junction between the middle and distal thirds of the segment, where the superficial femoral artery exits through the adductor hiatus. Typically, an initial stenosis progresses to occlusion; this is followed by proximal propagation of thrombus to the origin of the superficial femoral artery; outflow from the common femoral is then solely into the profunda femoris artery. Although distal propagation of thrombus may occur, the popliteal artery usually stays patent because stasis here is prevented by inflow from the profunda femoris via the geniculate collaterals. There is, therefore, often a patent vessel which can accept the distal end of a bypass graft.

Distal obliterative disease which is not associated with proximal atheroma should raise the suspicion of another pathology, especially diabetes mellitus, arteritis, or previous embolic disease. However the most common distal disease is found in older patients and is accompanied by extensive proximal atheroma. Distal disease is not amenable to reconstructive surgery.

Medical treatment may be indicated when the disease is not of sufficient severity to warrant operation (including angioplasty); when operation is impossible, inappropriate, or unsuccessful; or as an adjunct to operation. Several general measures are applicable to all patients whether or not they have surgery, for instance weight reduction in the obese, correction of anaemia or polycythaemia, treatment of hyperlipidaemia, and control of diabetes. The judicious treatment of heart failure and hypertension may also improve perfusion, but &bgr;-blocking drugs should be avoided as they may further compromise a diseased peripheral circulation.

Smoking is the most important correctable risk factor. Stopping smoking may be the only treatment that many patients require; claudication not infrequently improves spontaneously. Stopping smoking may not reduce atheroma that is already present, but continuation of smoking leads to an increased deposition and compromises the development of a collateral circulation. Smoking increases the risk of amputation and the incidence of graft occlusion after surgery.

Exercise is the other arm of effective conservation treatment: it may double the distance that can be walked before pain occurs in up to 80 per cent of patients. It has been suggested that selective exercise of those muscles which are most ischaemic produces the best results. Even rest pain may benefit from exercise, and it is prudent to recommend that patients exercise to the limit of comfort.

Apart from the above general measures, no other form of conservative treatment is widely accepted as likely to offer significant benefit in claudication or rest pain. A variety of drugs may occasionally offer a modest benefit; the most widely used are oxypentifylline (pentoxifylline) which alters red cell deformability so helping flow in the microcirculation, and naftidrofuryl which has an effect on ischaemic cell metabolism. More recently prostacyclin analogues have been used for their antithrombotic effect.

In a patient with unilateral symptoms it is not unusual to find angiographic evidence of early disease on the contralateral side. It is not yet known whether intervention for early asymptomatic disease confers benefit over a conservative policy and intervention in this group should occur only in the confines of a clinical trial addressing this question.

Intermittent claudication represents the ‘middle ground’. Any decision to intervene must take into account the possibility that spontaneous improvement may occur, especially if smoking is stopped and exercise adopted. Symptomatic improvement is especially likely within the first 6 months after onset of claudication. Consideration should be given to the degree to which the patient's lifestyle is affected and the hazard to life and limb that the proposed intervention might pose.

Rest pain or critical ischaemia despite appropriate medical management requires intervention if at all possible. Revascularization should generally be attempted if angiography and/or pulse generated run-off indicate that percutaneous transluminal angioplasty or reconstructive surgery is feasible. Nevertheless, in individual cases the chance of success may be recognizably slim and the decision to proceed will require fine judgement. There is a small group of patients, often unfit, who will benefit from primary amputation.

Percutaneous transluminal angioplasty is a radiological technique in which a guidewire is introduced percutaneously through the common femoral artery to lie within a stenosis. A catheter with a balloon at its end is introduced over the guidewire and the balloon is inflated within the narrowed segment (Fig. 5) 251. Although introduced by Dotter and Judkins in 1964 for the treatment of atheromatous femoropopliteal disease, percutaneous transluminal angioplasty was made popular primarily by Gruntzig and his colleagues. Today the technique is regularly applied to the femoropopliteal segment and to many sites other than the leg (Fig. 6) 252. It is carried out under local anaesthesia and is safe in experienced hands: the main complications of bleeding or thrombosis are unusual and can usually be remedied with surgical assistance. Percutaneous transluminal angioplasty is especially attractive for patients at the ends of the spectrum of disease severity, that is, those with minor claudication and for debilitated patients with severe ischaemia. Generally, occlusions up to 10 cm long are satisfactorily dealt with by this method.

It is sometimes impossible to push the guidewire and balloon catheter through an occlusion, especially if it is long. This may be overcome in some patients by burning through the occlusion using a laser (laser angioplasty). In general, the laser is not used to remove the bulk of atheroma but rather to allow proper placement of a guidewire and balloon catheter for conventional treatment. Laser angioplasty is a developing area and at this time is largely confined to specialist centres.

Mechanical atherectomy devices which may be introduced into the artery in a similar fashion to the angioplasty catheter are also available. The atherectomy catheter, however, has a cutting mechanism at its tip—a hollow core with side orifice and a sliding blade is a popular pattern, as is a high speed rotating cutter. These devices are used to cut through atheroma, retaining the resulting debris within their core for later removal. Their use is confined as yet to a few specialist centres.

The long-term results of laser angioplasty and atherectomy have yet to be assessed. They provide some hope for the use of minimally invasive procedures in the correction of arterial disease in the future and, if combined with a percutaneously inserted angioscope, may yet allow very sophisticated percutaneous arterial intervention.

When an acute occlusion occurs in a chronically atheromatous artery (usually the superficial femoral artery) it may cause limb threatening ischaemia for which emergency surgery used to be the only remedy. Today it is possible, in selected patients, to pass a long catheter percutaneously, via the common femoral artery, into the recent occlusion and to infuse a thrombolytic agent such as streptokinase or tissue plasminogen activator directly into the thrombus. If thrombolysis is successful the narrowed arterial segment may be improved by percutaneous transluminal angioplasty or selective operation.

The choice of operation for patients with occlusive disease of the lower limb arteries depends on the site of the occlusion(s), the availability of a suitable graft, and the experience of the operator. Although a variety of local bypasses or patch angioplasties may occasionally be desirable, by far the most common procedure is a bypass from the common femoral artery to a distal vessel; this is usually the popliteal artery—either above or below knee level—but it may also be to the tibioperoneal trunk or to any of the three (crural) vessels of the lower leg. The modern technique differs little from that first described by Kunlin in 1948; a shunt is constructed in parallel with the occluded artery using end-to-side anastomoses both proximally and distally (Fig. 7) 253. The rationale is to transport blood around an occluded segment while avoiding operative trauma to collaterals.

When available, the autogenous long saphenous vein is the best graft material for femorodistal bypass (Fig. 8) 255. However, this vein may be too small in calibre, thrombosed, markedly varicose, or may have been removed surgically in the past. Also, its usable length may be too short for the proposed operation. The short saphenous, the cephalic, and the basilic veins may then be used, individually or in combination with themselves or with a limited length of long saphenous vein. It is common for the vein to be assessed visually at the time of operation, but it is possible to assess the usefulness of the long saphenous vein before surgery, either by duplex ultrasound scanning or by saphenography.

If the vein is inadequate it may be necessary to use a graft of synthetic material, the most popular of which is expanded polytetrafluoroethylene (PTFE) (Fig. 9) 254. This inert substance has considerable resistance to thrombosis. Early patency rates for femoropopliteal bypass, both above and below knee, using such grafts are similar to those for autogenous vein (Fig. 10) 256. Late patency rates are better when vein is used, especially at the below-knee site. Nevertheless, late graft occlusion is not always associated with limb loss, even when the original operation was undertaken for critical ischaemia, and limb salvage rates for vein and polytetrafluoroethylene femoropopliteal bypasses remain broadly similar.

An alternative to PTFE is glutaraldehyde-stabilized human umbilical vein supported by an external polyester mesh. This tanned graft is non-antigenic and resists biodegradation reasonably well. Patency rates compare favourably with those of PTFE but umbilical vein is expensive and aneurysmal degeneration of the graft has been described. Other less popular grafts are available, such as those made from externally supported Dacron velour or ovine collagen.

When a graft of any type is inserted it is good practice to ensure at the end of the operation that the anastomoses, especially the distal one, are technically satisfactory and that flow through the graft is adequate to maintain patency. Several techniques are available. The electromagnetic flowmeter has been popular but it is difficult to achieve consistent results with this technique and peroperative Doppler ultrasonography is more satisfactory. To assess the integrity of an anastomosis the choice is between peroperative angiography (Fig. 11) 257 which is cheap and readily available, and angioscopy which is now becoming increasingly popular in major vascular centres (Fig. 12) 258,259. At the present time the authors' choice would be to use Doppler ultrasonography to assess the haemodynamics of the situation and angioscopy to assess the anastomoses. Nevertheless, if these techniques are not available it should not dissuade the surgeon from operating, especially in patients requiring limb salvage.

The most common reconstructive operation for occlusive disease below the inguinal ligament is a bypass from the common femoral to the popliteal artery, performed for occlusion of the superficial femoral artery. A bypass to the popliteal artery above the knee is the preferred option; if the above knee popliteal artery is very atheromatous or occluded then the below-knee vessel or even the crural vessels may be used.

Femoropopliteal bypass using reversed autogenous long saphenous vein is the archetypal operation. The popliteal artery is exposed via a medial incision and a satisfactory site for the lower anastomosis established. The femoral artery is exposed at the groin. A tunnel beneath sartorius is made between groin and suprageniculate politeal fossa, running orthotopically behind the knee if the infrageniculate site is to be reached. An adequate length of long saphenous vein is excised after tying and dividing its tributaries (Fig. 8) 255 and heparin is administered to the patient. It is generally agreed that for use as a reversed femoropopliteal bypass graft, the vein must have a minimum diameter of at least 4 mm. The vein is checked for leaks and if satisfactory is reversed to deactivate the valves, and anastomosed in an end to side fashion to the arteries.

Either the proximal or the distal anastomosis may be made first, but the inexperienced surgeon should complete the distal anastomosis before the proximal as this allows the leg to be fully extended with the graft lying in the subsartorial tunnel after one anastomosis in such a way that the exact length of graft which is needed is easily recognized. On completion of both anastomoses the bypass may be checked by assessing the lower anastomosis, either by angiogram or angioscopy via an untied vein tributary which is tied after the check procedure. Adequacy of flow through the graft may be checked using an electromagnetic flow meter or Doppler ultrasonography. In general, there is no need to reverse the heparin at the end of the procedure, nor is there any need for external drains.

This previously unfashionable operation has enjoyed renewed popularity since the late 1980s, possibly because better instruments for disrupting venous valves have become available. The concept is the same as for the reversed vein operation inasmuch as the sites for proximal and distal anastomoses are the same; the essential difference is that the vein is not excised but left in situ. The vein tributaries must all be recognized and tied off to prevent development of significant arteriovenous fistulae which compromise the graft blood flow distally, cause generalized oedema, and contribute to graft failure. The tributaries may be recognized by a variety of techniques from exposing the whole length of the vein for visual inspection (the authors' preference), to methods using Doppler ultrasonography, angiography, or angioscopy.

When using the vein in situ it is best to make the proximal anastomosis first and to declamp and allow the vein to fill with blood down to the first valve (Fig. 13) 260. A valve cutter (valvulotome) is then introduced from the distal end of the vein up to the femoral anastomosis (Fig. 14) 261,262. Withdrawing the valvulotome disrupts the valves and allows the blood to fill the graft as each set of valves is broken. Eventually the cutter is withdrawn followed by a spurt of pulsatile blood. The graft is then clamped and the lower anastomosis completed.

The major advantage of the in-situ technique, in addition to the fact that the vein and its blood supply is left largely undisturbed, is that the vein with the greater calibre in the groin is anastomosed to the large arteries, while the distal vein of smaller calibre is used in the anastomosis to smaller vessels further down the leg. It is therefore haemodynamically attractive and it is possible to achieve good results with a vein diameter of 3 mm or even less. Such grafts may be successful not only when anastomosed to the popliteal and proximal crural vessels, but even when the distal anastomosis is fashioned at ankle level. The technique, however, also has some technical disadvantages. Firstly, arteriovenous fistulae must be recognized and dealt with carefully. Secondly, the distal anastomosis may need to be made to a very small vessel and experience is required if this is to be successful. Such surgery is best done under magnification (×2.4 is adequate) and the result must be checked at completion by a peroperative angiogram or by angioscopy (if a thin enough scope is available).

Autogenous ipsilateral long saphenous vein should be used for bypass from the common femoral artery to levels below the knee joint if possible. Some surgeons believe that the superiority of vein is such that if the long saphenous vein is compromised by severe varicosity, thrombosis, or small calibre, a search should be made for another vein source, such as the contralateral long saphenous, the short saphenous, cephalic, or brachial veins. These may be used in combination as a vein–vein composite graft. Many others believe that if the ipsilateral long saphenous vein is unavailable it is reasonable to use a manufactured alternative. The most commonly used are made from PTFE.

PTFE grafts come in a variety of calibres but 6 mm (occasionally 8 mm) is usually chosen for bypasses in the leg (Fig. 9) 254. Some grafts are supported by external rings: these are an attractive adjunct, especially when the knee joint has to be crossed. The neointimal hyperplasia which occurs, especially at the distal anastomosis in PTFE grafts, has long been held to be the principle reason for their poorer patency (compared to vein) and it is likely that this tissue build-up is due to a compliance mismatch between the non-elastic graft and the expansile artery. The interposition of a piece of vein as a collar (Miller collar) between the PTFE and the artery has been suggested to be beneficial, but the results of meaningful trials of this promising method are still to be published.

In the presence of occlusive disease affecting the superficial femoral artery, the profunda femoris artery is the chief collateral channel between the iliac and popliteal arterial systems. In such circumstances, if the profunda is itself compromised, distal ischaemia is increased. Atheroma in the profunda shows a predilection for the area near its origin and reconstruction can be achieved by endarterectomy, by patch angioplasty (Fig. 15) 263 or by a combination of both. Such surgery may be carried out in association with inflow reconstruction, such as an aortofemoral bypass in which a graft limb is extended beyond the common femoral artery into the profunda femoris. In general a femorodistal bypass, if feasible, will produce a better result than a profunda reconstruction alone.

Apart from the profunda femoris artery, the only other artery in the leg which is at all frequently managed by endarterectomy is the common femoral. Likewise, patch angioplasty is rarely used at other native artery sites, although it has become fashionable for the relief of stenosis at graft–artery anastomoses and for the treatment of vein graft strictures.

The first lumbar sympathectomy was performed in 1924 and the operation has had variable popularity ever since. The advent of reconstructive surgery has put sympathectomy firmly in second place as a method of treatment of occlusive disease and at present its use can only be recommended as a last resort, for severe ischaemia when arterial bypass or reconstruction is not possible.

If either the collateral or the microcirculation is diseased, sympathectomy is unlikely to be of benefit. Its aim is to excise via a retroperitoneal route the sympathetic chain and ganglia from L4 to L2 (sometimes including L1), thereby increasing blood flow in the limb by decreasing vasomotor tone. The mechanism and effects of sympathectomy have been the subject of discussion for years, and they are complex; it is likely, however, that in suitable candidates skin capillary blood flow is increased. In contrast, the effect on skeletal muscle blood flow is controversial. Whatever the physiological effect, sympathectomy is extremely unlikely to benefit patients with claudication.

In the authors' experience, operative sympathectomy is only occasionally useful. Rather than submit patients to open surgery it may be more reasonable to perform a chemical sympathectomy by injecting 5 ml of 1:15 aqueous phenol at two separate sites along the lumbar sympathetic chain (Fig. 16) 264. This requires only local anaesthetic and for many years was performed ‘blind’, although today the placement of needles should be monitored radiologically. In experienced hands a chemical sympathectomy is as effective as an operative procedure.

This may be due to inadequate inflow or, more likely, outflow which has not been recognized before surgery. It may also be due to a technically poor anastomosis, especially at the distal end of the graft. A balloon catheter thrombectomy is only likely to be successful in the long term if the underlying cause of the thrombosis is also corrected.

This is usually due to technical misadventure and the insertion of further sutures into an anastomosis may occasionally be needed. Very rarely it may herald an early anastomotic infection.

This typically occurs from groin wounds, although wounds placed more distally in the leg are not immune. Most leaks stop spontaneously. Underlying grafts should be protected from the entry of external bacteria by administration of a broad spectrum antibiotic for the duration of leakage. Very rarely surgical closure of leaking lymph channels is necessary.

This is not uncommon after bypass surgery in the leg and is often due to cell swelling and an increase in interstitial fluid in reperfused critically ischaemic limbs. Oedema may also be due to lymphatic hold-up, which may be expected to subside rapidly without active intervention. Finally, leg swelling may be due to relative immobility after surgery or to deep vein thrombosis.

Wound infection is uncommon and occurs most commonly in the groin wound. It is usually superficial. The underlying graft should be protected by administration of an appropriate antibiotic. If the anastomosis or graft becomes infected the material used for reconstruction will generally need to be removed and a new reconstruction, by an alternative clean route, attempted. This may be difficult to achieve.

This is the main cause of late graft failure and is generally due to progression of atheromatous disease,either proximally or distally. It may also be due to a cellular build-up at the anastomotic site of the graft (neointimal hyperplasia). Strictures may form in vein grafts and compromise flow, finally causing thrombosis. Such strictures are most common within the first year after graft insertion and should be recognized early by regular follow-up so that they may be repaired before thrombosis occurs. This requires either percutaneous transluminal angioplasty or a patch graft.

When a leak occurs at an anastomosis the fluid may remain within the tissues and form a compartment contiguous with the artery/graft. Blood may flow freely in such a compartment and blood pressure will increase its size with time, forming a false aneurysm. An anastomotic aneurysm is a true aneurysm—it has an intimal lining—and reflects failing strength of the vessel wall at the site of anastomosis. Such aneurysms are more common after infection and may themselves be infected. They are also more common in patients with aneurysmal disease at other sites. Insertion of a graft which is too short, making it pull needlessly on its attached vessels, is a contributory factor in some cases.

A good quality autogenous saphenous vein of adequate calibre should rarely degenerate when used as an arterial conduit. Similarly, PTFE has been used in the leg for many years without major problems; this is probably true also for Dacron prostheses. The human umbilical vein graft may be prone to aneurysmal degeneration and other newer grafts—especially those of animal origin—must be regarded with caution at this time.

Patients who have undergone angioplasty or reconstruction should be subjected to regular follow-up. Occlusion of grafts or angioplasty sites is common and most likely to occur within a year. Although early graft occlusion—within 30 days of insertion—is usually due to technical error and will occur before graft surveillance has started, intimal hyperplasia and fibrotic stricture may be identified in the medium term prior to occlusion. Patency may be prolonged and the life of a graft extended if problems can be identified and corrected by percutaneous transluminal angioplasty or surgery before occlusion; correction after occlusion has occurred is far less likely to be successful.

The form which surveillance should take has recently been widely discussed. Simple clinical review of symptoms and pulses will not give adequate warning of an impending occlusion. Serial assessment of the ankle/brachial pressure index is easy, non-invasive, and is an effective screening test; a decrease in the index of more than 0.15 suggests that a stenosis is developing. However, even this test will fail to identify graft stenosis before occlusion in some patients. Angiography, especially using digital subtraction techniques, has been suggested as the most effective form of screening, but is both invasive and expensive. Duplex ultrasonography is cheaper and gives an adequate non-invasive assessment of graft function. If the scan shows a suspicious area, particularly a graft segment where the velocity of flow is very high, angiography can be used to confirm the abnormality and allow planning of corrective treatment. The best follow-up at this time entails a duplex scan at 6 and 12 weeks after intervention and then at 3-month intervals to 1 year. The gain from follow-up beyond 1 year is likely to be limited.

Although graft surveillance is important it does not replace the traditional management of a vascular graft. This involves stopping the patient smoking, which significantly improves graft patency, correcting polycythaemia and hyperlipidaemia, and treating diabetes mellitus. Most surgeons do not routinely prescribe long-term antiplatelet or anticoagulant drugs as the evidence that they benefit graft patency is small. The exception perhaps is low dose aspirin (150 or 300 mg/day) for patients with fabricated grafts below the inguinal ligament.

Adar R, Critchfield GC, Eddy DM. A confidence profile analysis of the results of femoropopliteal percutaneous transluminal angioplasty in the treatment of lower extremity ischaemia. J Vasc Surg 1989; 10: 57–67.

Ahn SS, et al. Removal of atheromatous lesions by angioscopically guided high-speed rotatory atherectomy. J Vasc Surg 1988; 7: 292–300.

Beard JD, et al. Pulse generated run-off (PGR): a new method of assessing calf vessel patency. Br J Surg 1988; 75: 361–3.

Beard JD, et al. Operative assessment of femorodistal grafts using a new Doppler flow meter. Br J Surg 1989; 76: 925–8.

Beck AH, et al. Long term results of percutaneous transluminal angioplasty: a study of 4750 dilatations and local lyses. Eur J Vasc Surg 1989; 3: 245–52.

Becker GJ, Katzen BT, Drake MD. Non-coronary angioplasty. Radiology 1989; 170: 921–40.

Bernstein EF, Fronek A. Current status of noninvasive tests in the diagnosis of peripheral arterial disease. Surg Clin North Am 1982; 62: 473–87.

Berridge DC, et al. Tissue plasminogen activator in peripheral arterial thrombolysis. Br J Surg 1990; 77: 179–82.

Bloor K. Natural history of arteriosclerosis of the lower extremities. Ann R Coll Surg Engl 1961; 28: 36–8.

Brewster DC, et al. Femoropopliteal graft failures: clinical consequences and success of secondary reconstruction. Arch Surg 1983; 118: 1043–7.

Clark, AM, et al. Anastomotic aneurysms of the femoral artery: aetiology and treatment. Br J Surg 1989; 76: 1014–7.

Cotton LT, Cross FW. Lumbar sympathectomy for arterial disease. Br J Surg 1985; 72: 678–83.

Dietrich EB, Timbadia E, Bahadir I. Application and limitations of laser assisted angioplasty. Eur J Vasc Surg 1989; 3: 61–70.

Dormandy J. European consensus on critical limb ischaemia. Lancet 1989; 334: 737–8.

Dormandy JA, Mahir MS. The natural history of peripheral atheromatous disease of legs. In: Greenhalgh RM, Jamieson CW, Nicolaides AN, eds. Vascular Surgery: Issues in Current Practice. New York: Grune & Stratton 1986: 3–17.

Dotter CT, Judkins MP. Transluminal treatment of arteriosclerotic obstruction. Description of a new technic and a preliminary report of its application. Circulation 1964; 30: 654–70.

Dunlop MG, et al. Vacuum drainage of groin wounds after vascular surgery: a controlled trial. Br J Surg 1990; 77: 562–3.

Ernst E. Peripheral vascular disease. Br Med J 1987; 299: 873.

Ernst E, Matrai A. Intermittent claudication, exercise and blood rheology. Circulation 1987; 76: 1110–4.

Green RM, Roedersheimer LR, DeWeese JA. Effects of aspirin and dipyridamole on expanded polytetrafluoroethylene graft patency. Surgery 1982; 92: 1016–26.

Grigg MJ, Nicolaides AN, Wolfe JHN. Detection and grading of femoro-distal graft stenoses: duplex velocity measurements compared with angiography. J Vasc Surg 1988; 8: 661–6.

Gruntzig AR, Hopff H. Perkutane Rekanalisation chronischer arterieller Verschlusse mit einemneuen Dilatationskatheter. Deutsch Med Wochenschr 1974; 99: 2502–5.

Haimovici H, Callow AD, De Palma RG, Ernst CB, Hollier LH, eds. Vascular Surgery. 3rd edn. Norwalk, Connecticut: Appleton and Large.

Harris PL, How TV, Jones DR. Prospectively randomised clinical trial to compare in situ and reversed saphenous vein grafts for femoropopliteal bypass. Br J Surg 1987; 74: 252–5.

Hess H, Mietaschk A, Brucki R. Peripheral arterial occlusions: a six year experience with local low dose thrombolytic therapy. Radiology 1987; 163: 753–8.

Hughson WG, et al. Intermittent claudication: factors determining outcome. Br Med J 1978; i: 1377–9.

Jelnes R, et al. Fate in intermittent claudication: outcome and risk factors. Br Med J 1986; 293: 1137–40.

Juergens JL, Barker NW, Hines EA. Arteriosclerosis obliterans: a review of 520 cases with special reference to pathogenic factors and prognosis. Circulation 1960; 21: 188–95.

Kakkasseril JS, et al. Efficacy of low dose streptokinase in acute arterial occlusion and graft thrombosis. Arch Surg 1985; 120: 427–9.

Kannel WB, Shurtleff D. The Framingham study: cigarettes and the development of intermittent claudication. Geriatrics 1973; 28: 61–8.

Kunlin J. Le traitemente de lárterite obliterante par la greffe veineuse. Arch Mal Coeur 1949; 42: 371.

Leopold PW, et al. Role of B-mode venous mapping in infrainguinal in situ vein—arterial bypasses. Br J Surg 1989; 76: 305–7.

Michaels JA. Choice of material for above-knee femorodistal bypass graft. Br J Surg 1989; 76: 7–14.

Miyata T, et al. A clinicopathological study of aneurysm formation of glutaraldehyde–tanned human umbilical vein grafts. J Vasc Surg 1989; 10: 605–11.

Moody P, et al. Asymptomatic strictures in femoro-popliteal vein grafts. Eur J Vasc Surg 1989; 3: 389–92.

Moody P, Gould DA, Harris PL. Vein graft surveillance improves patency in femoro-popliteal bypass. Eur J Vasc Surg 1990; 4: 117–21.

Negus D, Irving JD, Friedgood A. Intra-arterial prostacyclin compared to Praxilene in the management of severe lower limb ischaemia: a double-blind trial. J Cardiovasc Surg (Torino) 1987; 28: 196–9.

Pond GD, et al. Digital subtraction angiography of peripheral vascular bypass procedures. Am J Roentgenol 1982; 138: 279–81.

Prevention of atherosclerotic complications with ketanserin trial group. Prevention of atherosclerotic complications: controlled trial of ketanserin. Br Med J 1989; 298: 424–30.

Quick CRG, Cotton LT. The measured effect of stopping smoking on intermittent claudication. Br J Surg 1982; 69(suppl): 24–6.

Rutherford RB, ed. Vascular Surgery. 3rd edn. Philadelphia; W. B. Saunders: 1989.

Schubotz R, Muhlfellner O. The effect of pentoxifylline on erythrocyte deformability and on phosphatide fatty acid distribution in the erythrocyte membrane. Curr Med Res Opin 1977; 4: 609–17.

Shaw SWJ, Johnson RH. The effect of naftidrofuryl on the metabolic response to exercise in man. Acta Neurol Scand 1975; 52: 231–7.

Simpson JB, et al. Transluminal atherectomy for occlusive peripheral vascular disease. Am J Cardiol 1987; 61: 97–101.

Skagseth E, Hall KV. In situ vein bypasses: experiences with the new vein valve strippers. Scand J Thor Cardiovasc Surg 1973; 7: 53–8.

Strandness DE, Sumner DS. The relationship between calf blood flow and ankle blood pressure in patients with intermittent claudication. Surgery 1969; 65: 763–6.

Thompson JF, et al. Intervention for graft stenosis: the role of surgery and transluminal angioplasty. Br J Surg 1989; 76: 1017.

Towne JB. Role of fibrointimal hyperplasia in vein graft failure. J Vasc Surg 1989; 10: 583–4.

Tyrrell MR, et al. Experimental evidence to support the use of interposition vein collars/patches in distal PTFE anastomosis. Eur J Vasc Surg 1990; 4: 95–101.

Veith F, et al. Preoperative saphenous venography in arterial reconstructive surgery of the lower limb. Surgery 1979; 83: 253–6.

Wiseman S, et al. The influence of smoking and plasma factors on prosthetic graft patency. Eur J Vasc Surg 1990; 4: 57–61.

White GH, et al. Intraoperative video angioscopy compared with angiography during peripheral vascular operations. J Vasc Surg 1987; 6: 488–95.

White RA, White GH. Laser angioplasty: development, current status, and future perspectives. Sem Vasc Surg 1989; 2: 123–42.

Whittemore AD, et al. What is the proper role of polytetrafluoroethylene grafts in infrainguinal reconstruction. J Vasc Surg 1989; 10: 299–305.

Wiseman S, et al. The influence of smoking and plasma factors on prosthetic graft patency. Eur J Vasc Surg 1990; 4: 57–61.

Woelf KD, et al. Intraoperative assessment of in situ saphenous vein bypass grafts by vascular endoscopy. Eur J Vasc Surg 1988; 2: 257–62.

Wolfe JHN, McPherson GAD. The failing femoro-distal graft. Eur J Vasc Surg 1987; 1: 295–6.