• type A lesions should usually be treated by endovascular means;

• type B lesions should preferentially be treated by endovascular means;

• type C lesions should preferentially be treated by open revascularisation;

• type D lesions should usually be treated by open surgery.

Abdominal aorta

Haemodynamically significant stenoses of the infrarenal aorta are rare and usually seen in short, obese female smokers with a small-calibre aorta. Prior to the advent of endovascular therapies these patients would have undergone a localised aortic endarterectomy but primary stenting is now the first-line treatment (Fig. 4.2), with 5-year patency rates of 50%.¹² Complications are rare and usually relate to embolisation, pseudoaneurysm formation and recurrent disease.¹³ Complete aortic occlusions are best treated by an aorto-bifemoral bypass (see later). In an unfit patient aorto-uni-iliac stenting, combined with a femorofemoral crossover, may be a better alternative to an axillo-bifemoral bypass.¹⁴ Kissing stents can be used but there is a risk of one stent compressing the other.

Iliac arteries

Though the European guidelines on the management of CLI and the diabetic foot¹⁸ state that there are no randomised studies of percutaneous transluminal angioplasty (PTA) to open surgery in the iliac arteries, this is incorrect.

There is one RCT that compares open surgery to angioplasty in symptomatic iliac disease.¹⁹ In this study, 263 men with iliac disease that was a major component to either rest pain or lifestyle-limiting claudication underwent either bypass surgery (n = 123) or angioplasty (n = 129). There were three deaths in the surgery group and none in the angioplasty group. There was no difference in patency and limb salvage rates, at a median follow-up of 4 years. The authors concluded that the lower morbidity and mortality rate in the angioplasty group supported the concept of angioplasty first strategy as outcomes between the two treatments were no different. The study was, however, limited by (i) the absence of women, who have smaller arteries and (ii) the relatively small numbers, which precluded subgroup analysis.

In a second, non-randomised comparative study the complication rate, primary patency and cost of stent deployment were compared with direct surgical reconstruction for the treatment of severe aorto-iliac occlusive disease.²⁰ Sixty-five patients underwent stent deployment and 54 patients had surgical reconstruction. No significant difference was observed between the groups in terms of clinical presentation, demographics and late complications. The cumulative primary patency rate for bypass grafts was, however, significantly better than for iliac stents at 18 months (93% vs. 73%), 30 months (93% vs. 68%) and 42 months (93% vs. 68%). Multivariate analysis suggested that females, anyone with ipsilateral superficial femoral artery (SFA) occlusion, and patients who had procedure-related vascular complications and hypercholesterolaemia were more likely to thrombose their bypass graft or stent. Costs did not differ significantly.

So here we have two studies with very different conclusions. The first is an RCT and the second an observational study. It is difficult to determine which of the many variables in the observational study affected which group, and it is of course possible that patients who were less fit for open surgery were treated by endovascular means. The fact is that the literature remains controversial.

When is an iliac lesion significant?

The arteriographic grading of stenoses is hampered by wide intra- and inter-observer variability.²¹ Intravascular ultrasound provides better measurements of cross-sectional area but it is expensive and not widely available. In theory, pressure gradient measurements at the time of angiography should be a relatively simple procedure. However, there is no consensus in the literature regarding what constitutes a significant pressure gradient²² and there is little or no information on the role of vasodilators (doses), systolic versus mean pressures and peripheral resistance in assessing stenotic vessels. Tetteroo et al. reported that they would have stented between 4% and 87% of their patients, emphasising the overall lack of consensus.²³ They suggested that a mean gradient of 10 mmHg at rest or after vasodilation should be used as a marker of significance and this is generally accepted, though there is no evidence at all for this.

Who should have an angioplasty or a stent?

Early aorto-iliac angioplasty had a high rate of embolic complications, particularly when dealing with occlusive rather than stenotic disease.²⁴ As a result stenting became the first-line treatment of iliac occlusive disease based on this observation. RCTs have been performed, but these have been compounded by flawed design and study methodology. Such a study demonstrated 4-year patency rates of 91.6% for stents and 74.3% for PTA alone, but the results have never been fully published.²⁵

There is only one peer-reviewed, published RCT that compares primary stent placement with PTA followed by stent placement if needed in the iliac arteries. The Dutch Iliac Stent Trial (DIST)²⁶ randomised patients to either a primary Palmaz stent or PTA alone. In the latter group stents were reserved for patients with suboptimal PTA results. The authors reported that 43% of patients randomised to PTA received iliac stents and that 2-year cumulative patency rates for the two groups were similar at 71% versus 70%. They concluded that because angioplasty followed by selective stent placement is less expensive than primary stent placement, the former should be the treatment of choice for lifestyle-limiting intermittent claudication caused by iliac artery occlusive disease. However, only 29 iliac artery occlusions were treated among 279 patients and, by study design, patients with occlusions more than 5 cm in length were ineligible for enrolment. PTA alone failed in 10 out of 12 subjects (83%). Although stents seem to offer special benefits in the treatment of longer segment lesions, little guidance about such patients is provided by this trial. Furthermore, most patients in the DIST study had mild clinical symptoms, and only 22% of patients in each group were classified as having Society for Vascular Surgery/International Society for Cardiovascular Surgery (SVS/ISCVS) grade 3–5 ischaemia. As a consequence, the milder pattern of atherosclerotic disease observed in the DIST study differed from that encountered in many interventional practices. In addition, the study was designed to have a 90% likelihood of detecting a 10% difference in post-treatment arterial patency after 12 months between the two groups (power = 90%). Unfortunately, the trial was terminated after less than 80% of the intended sample had been enrolled. Outcome information after as little as 1 year is available in the published report for about 60% of subjects who ultimately enrolled, a fraction that actually represents less than 45% of the intended patient sample. Thus, even though the authors did not observe ‘substantial differences’ in clinical outcomes between their experimental groups, the DIST study was more likely to miss than to detect the same potential 10% improvement in 12-month patency rates that was used to configure the trial, as the actual power was less than 50%. Because of under-recruitment, the authors attempted to amplify their sample by reporting the number of lesions treated, rather than the number of subjects actually involved. For example, a patient with simultaneous common and external iliac stenoses or occlusions was classified as two treated lesions, rather than as a single person. As a result, little information about crucial patient subgroups (e.g. occlusions vs. stenoses, common iliac vs. external iliac lesions) can be gleaned from this report. Despite these problems the authors continued to publish from the original dataset, reporting patency rates and patient survival at 5 years.²⁷

There is a second RCT of stent versus angioplasty in complete iliac occlusions (the STAG trial). This is not yet published in the peer-reviewed literature but has been presented.²⁸ The results suggest that there are significantly fewer major complications, particularly embolisation, following stent compared to PTA, but that stents confer no benefit in terms of patency (P. Gaines, personal communication). If so, we are brought full circle to the original premise that stenting infers benefits in terms of reduced complications in iliac occlusions, rather than improved patency. Though the authors are not in a position to comment until the publication of this study, much will depend on the original design and power of the study in terms of its primary and secondary end-points.

Iliac stenoses

In iliac stenotic disease, angioplasty alone should be the procedure of choice. The end-point for aorto-iliac revascularisation procedures is determined by intra-arterial haemodynamic measurements, which are readily obtained in this arterial segment. Ideally, simultaneous waveforms obtained above and below the treated segment should overlap, with no mean or systolic gradient (Fig. 4.3).

Often, ideal haemodynamic results are not achieved because of recoil or patient intolerance to balloon inflation to a calibre consistent with elimination of the gradient. This situation may demand a stent of suitable diameter and length. Balloon-mounted stents are normally preferred in this situation, though self-expanding stents may be desirable if the problem is intolerance.

Iliac occlusions

Chronic iliac occlusions can be crossed from the ipsilateral or contralateral side and stented to reduce the risk of embolic complications. This requires the insertion of a stent prior to PTA (Fig. 4.4). Self-expanding stents are indicated and in the authors’ experience this will always traverse the chronic occlusion over the wire. Heparin is mandatory immediately after arterial access. Once a stent of appropriate diameter and length has been deployed it can then be inflated with a suitable sized balloon. If there is a suitable iliac stump of the ostium, a uni-iliac stent will work well. If, however, the occluded iliac artery does not have a stump and the aortic lumen tapers smoothly to the occlusion, then ‘kissing’ iliac stents may have to be used, even when the contralateral iliac artery does not have a significant stenosis.³¹ This should be avoided if at all possible. Early results are good³² but, in small arteries, the long-term patency is not as good as uni-iliac stenting. The external iliac artery is slightly more prone to rupture than the common iliac, but responds well to stenting.

Complications

The third British Iliac Angioplasty Study (BIAS)³³ followed 2233 patients for 12 months to 8 years. The complication rate in patients suffering intermittent claudication was 2.6% and in patients with critical leg ischaemia was 11.6%, reflecting the difference in age, comorbidity and disease extent. However, only 2.4% of patients overall required unplanned surgery, unplanned endovascuscular intervention or some other procedure that delayed discharge from hospital. The overall mortality for inpatient intervention was 2.8% and for day-case intervention no deaths were reported. Importantly, arterial perforation remains rare, with an overall mortality of 0.17%.

Potentially serious complications include the following:

Aorto-bifemoral bypass

The infrarenal abdominal aorta can be approached laparoscopically or by conventional open surgery. Laparoscopic aortic surgery has been introduced in many European countries and advocates claim lower morbidity and mortality rates and shorter hospital stays.^41,42 There remains no convincing evidence to support these claims as there have been no randomised trials. Laparoscopic aortic surgery requires extensive training and the procedure times are longer than for open surgery. With regard to open surgery, there is little or no evidence to favour either a transperitoneal versus a retroperitoneal approach or a vertical versus transverse incision, although transverse incisions are favoured by anaesthetists as the top end of a vertical incision often ‘escapes’ an epidural. End-to-end anastomoses are used for flush occlusions below the renal arteries (Fig. 4.7). If the proximal infrarenal aorta is patent, then an end-to-side anastomosis can preserve flow to a patent inferior mesenteric and/or internal iliac artery, thus reducing the risk of pelvic ischaemia.

Axillo-bifemoral bypass

Often described as the last resort of the destitute, this has its place in the very unfit patient and those with a ‘hostile’ abdomen.

Unilateral iliofemoral bypass

This procedure is indicated where there is extensive disease within the external iliac artery, which has failed to respond to endovascular intervention or where the disease extends into the common femoral artery.

The key to success is a good inflow in the form of a disease-free common iliac artery and appropriate run-off into the SFA or profunda femoris artery (PFA). In the majority of cases a common femoral artery (CFA) endarterectomy and patch will have to be performed to gain access to the run-off vessels. In the rare cases where the common iliac is poor, the distal aorta can be used as a source of inflow. The common iliac artery is approached through a classical Rutherford–Morrison incision and an ‘Omnitract’ or ring retractor is a valuable tool in maintaining exposure of the vessels. Care should be taken to identify and preserve the ureter. Proximal control of the common iliac artery is essential, and some surgeons would avoid dissecting around the back of the vessel, as torrential venous bleeding can ensue, especially if there has been an inflammatory response around the vessels. Some surgeons would therefore advocate not to fully mobilise the vessel, but merely clamp it, without slings. Previous abdominal surgery, e.g. appendicectomy, or inguinal hernia repair may affect access to the retroperitoneal space and should be considered prior to any planned surgery. Externally supported 8-mm Dacron or PTFE are the conduits of choice and are usually sewn end-to-side to the common iliac. A tunnel is then created, from above and below the inguinal ligament, taking care not to injure the vein that overlies the CFA at the level of the inguinal ligament and the distal anastomosis completed to the CFA. In those cases where the SFA is occluded it is essential to extend the anastomosis down to at least the first bifurcation of the PFA.

Iliofemoral crossover

This procedure is indicated where the affected limb has no inflow vessel but the contralateral limb has a relatively disease-free iliac inflow. Use of the contralateral iliac artery avoids exposure of both groins. In addition, if the donor iliac artery develops significant atherosclerosis over time, this can be treated percutaneously via the donor femoral artery, thus avoiding direct puncture of the graft and/or anastomotic site. The donor iliac vessels are approached via a Rutherford–Morrison incision and the graft (Dacron/PTFE externally supported 8 mm) is placed in a lazy’ S position. It is essential that the patient is catheterised and great care should be taken tunnelling the graft into the contralateral groin as there is the potential to damage the bladder and the iliofemoral veins.