Perioperative care after EVAR

This chapter belongs in 2.1.18 Peri-operative Issues in Intensive Care, where vascular surgery is listed as an L2 peri-operative topic. Whereas vascular surgery is vast and diverse, the ICU experience of vascular surgery is a small narrow window into their worst complications, least carefully chosen patients, and riskiest gambits. The focus of this ostensibly EVAR-focused chapter will be mostly on the detection and handling of the worst outcomes, via the reasoning that the intensivist is only of much use when things are going wrong, and that the routine post-op care of such patients is going to be handled mostly by protocolised pathways that do not require anything especially imaginative. Confirming this, Wong et al (2025) found that ICU admission for routine ICU patients had no effect on improving their mortality.  This also seems to reflect the examiner sentiment, as the EVAR questions from the past papers were generally interested in the management and prevention of disasters:

• Question 16 from the first paper of 2026 - ischaemic colitis post EVAR
• Question 3 from the second paper of 2025 - EVAR for ruptured AAA with day-2 bloody diarrhoea and suspected ischaemic colitis; mechanisms, investigations with expected findings, and management principles for bowel ischaemia after EVAR.
• Question 27 from the second paper of 2022 - Resuscitation of the septic shock patient.
• Question 29 from the second paper of 2019 - Causes of renal failure and timing of dialysis.
• Question 6.2 from the second paper of 2012 - Rhabdomyolysis following AAA repair.
• Question 24 from the first paper of 2007 - Causes of renal failure; also rhabdomyolysis as a cause of acute kidney injury.
• Question 29 from the first paper of 2005 - Prevention of acute renal failure (contrast-induced nephropathy) in AAA repair.

The most helpful papers for this would probably have to be De Paulis et al (2022), which covers the subject in broad brushstrokes but in a pleasing readable structure. More specifically to the complications are Daye & Walker (2018).  Surgical literature describing the indications for this is interesting, but probably less relevant, as the ICU are rarely involved in the decisionmaking around the type and timing of the intervention, and so the ESVS 2024 guidelines are mentioned here mostly for the reader to be aware of them, but not as a recommendation to immediately go and read them.

Acronymity among EVARs

Endovascular aortic procedures are severely acronymised, with numerous uppercase letter combinations to bedevil the reader of surgical notes. A selection includes:

• sEVAR: standard EVAR, usually conventional infrarenal EVAR.
• cEVAR: complex EVAR, this being fenestrated, branched, chimney, iliac branch, or where the access required cutdown
• fEVAR / FEVAR / F-EVAR: fenestrated EVAR, using fenestrations or scallops to preserve renal or visceral branch vessels.
• bEVAR / BEVAR / B-EVAR: branched EVAR, where branch grafts protect the aforementioned vessels instead of fenestrations
• f/bEVAR / F-BEVAR / FB-EVAR: fenestrated and branched EVAR, a term which catches combined devices.
• chEVAR / Ch-EVAR / CHEVAR: chimney EVAR, where a parallel covered stent beside the main aortic graft is used to preserve a branch vessel and extend the seal zone.
• Sn-EVAR / SnEVAR: snorkel EVAR, which is a term that seems to be interchangeable with chimney EVAR.
• PG-EVAR / PGEVAR: parallel graft EVAR, an umbrella term for chimney, snorkel, "periscope", and whatever other configuration of parallel tubeplay
• PEVAR / P-EVAR: percutaneous EVAR, performed through percutaneous femoral access with closure devices rather than surgical femoral cutdown, which is really just the standard version much of the time.
• rEVAR / REVAR / r-EVAR: ruptured EVAR, meaning EVAR performed for ruptured abdominal aortic aneurysm.
• eEVAR: emergency EVAR; or, ...elective EVAR in some datasets; so not an especially informative acronym
• iEVAR: intact EVAR, meaning non-ruptured; i.e. not an emergency.
• AUI EVAR: aorto-uni-iliac graft configuration, usually requiring contralateral iliac occlusion and femoro-femoral crossover grafting.
• TEVAR: thoracic endovascular aortic repair

It is neither reasonable to expect, nor possible to examine, the candidate's appreciation of all these variants, and this list is therefore left here mainly in protest against the proliferation of different names for the exact same thing.

Natural history of recovery after uncomplicated elective EVAR

For an elective infrarenal EVAR procedure that contained nothing memorable for the surgeon and anaesthetist, the normal recovery is short and unexciting. In a National Surgical Quality Improvement Program cohort of 11,229 elective EVAR patients, Gupta et al (2014) reported a median hospital stay of only 2 days. If something is going to go wrong, it seems to go wrong in the first six hours or so, according to Shaw et al (2019), so the early period of observation is the most important, and if you survive those early hours, you are probably going to be fine. Where patients are carefully selected and the volume is high (i.e. the teams are well-practiced) even same-day discharge can be contemplated, although the linked article is a case series of only four such cases. 

So, do these people even need ICU? Most people would argue that yes, they do, on the basis of the fact that the NELA report recommends routine ICU admission for any patients with a postoperative risk of death ≥5%; but in fact there may be little benefit to this, especially if advanced perioperative junior medical and nursing species have evolved in the local perioperative basin. Moreover many would argue that in the modern vascular surgical practice the 30-day risk of death from these procedures should be <1%. So if your EVAR patients are requiring ICU routinely, it must mean that the volume is not large, and the patients are medically and surgically challenging. That is often the kind of thing seen in major Australian tertiary public hospitals, as the less complex patients get funnelled to the peripheral centres and up the road to the private. 

In summary, the normal ERAS-era trajectory of an uncomplicated EVAR destined to achieve the promised 99.1%  30-day survival should be:

• Be a fairly comorbid person in their mid-70s, probably an overweight male
• Have a 2-3 hour procedure with GA
• Get extubated in theatre
• Recover from GA
• Transfer to ICU with strict instructions for supine bed rest (to protect the groin access site from haematoma); for this reason, the ability to lay flat  for a prolonged period without shortness of breath is an essential part of the preoperative screening process.
• Be reasinably pain-free during this period because usually some sort of regional technique has kept the groin site from hurting, and because the aorta is insensate to the presence of a stent inside it. 
  • This in itself is an issue, because the patient will not be aware of their haematoma, and so the groin becomes the object of much attention, after decades of neglect.
  • Limb neurological and vascular observations and arterial blood gases are frequent in this period, as the recovering patient may remain unaware of gut or muscle ischaemia in the wake of a spinal anaesthetic (or a spinal infarction)
• Get aspirin and a statin later on day 0, as well as a normal diet and DVT prophylaxis (the option of restarting normal oral diabetic medications is open, if the oral intake is enough to impress the evening nurses)
• Sit out of bed and mobilise early on the following day, or even later same day, with the objective of going home if mobile.

Deviations from this pathway represent some kind of abnormality and should be viewed as suspicious. It is not normal to have abdominal pain, metabolic acidosis, oliguria or severe nausea after these procedures. 

Complications following EVAR

The aorta being a large blood-filled thing of central importance makes the complications from its instrumentation early and obvious, so it stands to reason that the immediate post operative focus should be on the vascular access sites, embolic phenomena and the misbehaviour of the branches of the abdominal aorta. Maleux et al (2009) lists these in some detail:

• Endoleak (more on that later) - highest incidence, 10-50%
• Access site complications (1-10%, haematoma etc)
• Specifically, arterial thombosis, pseudoaneurysm or dissection in ~3%
• Contrast-induced nephropathy( the dye is being injected, effectively, right into the kidneys, and typically 50-100ml is used); ~6.7%

Having interfered with the major source of blood supply to the lower body in a patient with clearly abundant cardiovascular risk factors suggests the aorta's intimal surface may not be particularly clean, and the potential for embolic complications is clearly very high. 

The Things that Might Infarct:

• Bowel (~3%)
• Kidneys 
• Lumbar spine and paravertebral muscles
• Spinal cord (~0.2%)
• Testicles
• Legs and feet

Kidneys may also be affected by being accidentally occluded by the graft, a gaffe Maleux et al attribute to "a learning curve of the endovascular team". The rate given in the paper is an astonishing 5%, though the authors do acknowledge how old their reference for this is. Additionally, the limbs of the graft can migrate, kink or fracture, and more disturbingly this can happen at any time, not just immediately post op.

Which brings us to the discussion of late complications. Post-discharge risk is non trivial, with something like 40% of 30-day morbidities and 31% of 30-day deaths occurring after discharge. Late stent complications also include:

• thrombosis, kinking or migration of the stent
• embolic phenomena
• stent infection
• late endoleak
• boring routine complications of surgery like wound infection, VTE and pneumonia
• Non-boring complications like aortoenteric fistula

Endoleak

"Endoleak" is the delightful colloquialism to describe "continuing blood flow around the graft into the aneurysm sac", a term coined by White et al (1997) at RPAH in Australia which appears to have been accepted immediately and became the official terminology for this complication. It comes in several varieties, which can be first subdivided into being primary or secondary, where primary are those that occur within 30 days, and secondary are those happening later.

Endoleak type	Location of Leak	Incidence (%)
Type I	Attachment sites	2–10
A	Proximal end
B	Distal end
C	Iliac occluder
Type II	Retrograde flow through patent aortic side branches	8–29
A	Single vessel
B	Multiple vessels
Type III	Mechanical failure	1–5
A	Modular disconnection
B	Fabric tear
C	Junctional separation (fenestration, branch, visceral stent)
Type IV	Graft porosity	< 1
Type V	Aneurysm sac enlargement without visualised endoleak	2–3

Why are these bad? Well, for one, the underlying problem (AAA) remains unresolved, as the abdominal aorta still has some aneurysmal dilatation in it, which has a lot of blood still going into it through the endoleak. This means one has done nothing to reduce the risk of rupture, and potentially have produced turbulence that facilitates it. That this is the most common complication is concerning (few other surgeries have a 10-50% likelihood of being pointless). 

What do you do about them?

• Type I and III need to be handled sooner rather than later. The graft failed to seal the aneurysmal sack, either by remaining open at the top end (Type I) or by losing integrity and perforating (Type III, which are much more rare).  Now there is a jet of turbulent blood being injected into a smaller space than was previously available, plus now a large foreign body is rubbing against the walls.
• Type II can sometimes just be watched, and about 50% resolve spontaneously, but others with high risk features may need intervention.

It does not help that the aorta is relatively insensate to this sort of thing,  ad the endoleak may present no clinical features that can be detected externally, except perhaps a bruit.  Wooley et al (1998) reports that there are pain fibres in there (and patients undergoing balloon dilatation for aortic coarctation do report pain while the balloon is inflated), but for this to happen, one needs to be distorting the adventitia, and by the time you are doing that the aorta has already dissected or ruptured. 

Bowel ischaemia following EVAR

The exam candidate will be relieved that there is indeed something here that might relate to Question 3 from the second paper of 2025, where detailed questions about bowel ischaemia were the main theme. The generic background for this belongs in the chapter on mesenteric ischaemia, and the ensuing text only adds a perioperative flavour. Bowel ischaemia after AAA repair is uncommon after elective EVAR but is much more likely after rupture, shock and open repair; and when it occurs it is a high-mortality complication (Gurakar et al, 2019; Aday et al, 2018).

• Mechanisms of bowel ischaemia following EVAR:
  • Shock state
  • Abdominal compartment syndrome from haematoma
  • Reperfusion injury, after flow is restored
  • Occlusion or dissection of mesnteric vessels during stent procedure
  • Shower of emboli during stent deployment
  • Stent migration or limb kinking
  • Anatomical vulnerability at watershed areas
  • Pre-existing vulnerable supply (i.e the entire vascular tree is, like the aorta, covered in cheese)
  • It was already like that (i.e. the rupture of the aneurysm had already compromised the blood supply of the gut, and restoring it later with a stent has done nothing to reverse the injury)
• Investigations for bowel ischaemia following EVAR:
  • ABG  is tempting because lactate, but one must admit that this is a fairly indistinct signal, and a raised lactate could mean anything. But it is fast and it gives you a reason to worry.
  • End-organ function biomarkers like urea, creatinine, amylase and lipase, etc - to see what else has lost perfusion
  • Specific biomarkers of gut ischaemia are described in the literature, but are not in routine use. Promising ones (i.e. ones which you can actually order at your unit) are probably limited to D-lactate.
  • CT mesenteric angiogram would be the most effective way of assessing the aorta and the mesenteric circulation all at once, to determine whether there has been some stent malfunction or large-vessel compromise. 
  • DSA (angiography) would be an exploratory option which has the associated benefit of potentially leading to some kind of reperfusion solution, and could be the follow-up investigation if emboli or stent misbehaviour are identified on CT.
  • Endoscopy can confirm the viability of the bowel, and give some warning regarding whether a perforation is imminent. On the other hand, a theoretical pending perforation can transform into a real current perforation during colonoscopy. The expected early findings looks like some kind of generic colitis (inflamed angry looking bowel with areas of ulceration), progressing to grey and black as the bowel becomes gangrenous. 

Imaging findings 

The examiner's comments for Question 16 from the first paper of 2026 were insistent on the importance of being able to recognise the features of bowel ischaemia on CT, which seems important, but also seems like something that could have been a radiology viva. Olson et al (2023) describes these as follows:

• Acute arterial ischaemia:
  • Bowel wall thinning (“paper thin”)
  • Diminished or absent mural enhancement on arterial phase
  • Obviously, a large occluded vessel
• Acute venous ischaemia:
  • Mural thickening
  • Mural stratification (one should not usually be able to clearly see the layers of the bowel wall on CT)
  • Mesenteric stranding / oedema
• Generally,
  • Bowel dilatation
  • Pneumatosis coli, where the bowel wall has gas in it
  • Portal venous gas (distinguishable from gas in the biliary tree because it is seen in the periphery of the liver, whereas biliary gas is more central and perihilar).
  • Pneumoperitoneum of perforation

Endoscopy for mesenteric ischaemia following EVAR

The college also listed endoscopy as a part of the above-standard answer, which suggests that the ability to recognise this as an option is seen as an advanced stage of development. In fact, this is a scenario seen more and more often. Kim et al (2023) observed retrospectively their institutional practice of performing routine colonoscopy on all post-AAA-rupture emergency EVAR patients and found that 25% of them had some degree of colitis, and that it increased their mortality from 6% to 26%. For Lozano-Maya et al (2010), it changed management in almost 60%, mostly by revealing high grade ischaemia that motivated earlier surgery.  Specifically, the grades of ischaemia are this three-tier system dating back to Favier et al (1976):

• Stage 1 (Mucosal Ischemia): Characterized by mucosal erythema, petechial hemorrhages, and superficial ulcerations confined to the mucosal layer.
• Stage 2 (Transmural Extension): Demonstrates progression to deeper ulcerations penetrating the muscularis mucosae, with ischemic changes extending into the submucosal layer.
• Stage 3 (Full-thickness Necrosis): Presents with transmural necrosis involving the muscularis propria, or frank perforation.

Management for bowel ischaemia following EVAR 

Like with every management answer for a condition with a multitude of causes, this one has the capacity to paralyse the exam candidate with indecision, as the correct answer is obviously going to depend on the specific thing that happened, and a broad generic answer seems incompetent because it would not apply equally to every possibility. One will clearly do something different if the ischaemia is being caused by a thromoembolic occlusion, vs. stent malposition. However it seems the CICM examiners obviously wanted something haemodynamic to headline the response to Question 16 from the first paper of 2026. "Specific haemodynamic management for optimisation of GI blood flow" was identified as a key identifying feature of those candidates who gained more marks.

From this, one could reasonably infer that, if the question is sufficient for examination in the Fellowship, there must exist some recognised body of knowledge against which the candidate's attainments may properly be measured. Of course nothing could be further from the truth, which reflects the clinical realities of this specialty. Of the papers on this subject, some offer a single line of advice (Steele, 2007)  and others ignorare and one might successfully argue that it is in fact more important to test the knowledge and the judgment of the candidates on questions that do not have a well-accepted answer, because that is where the intensivist's command is properly tested. 

• Reperfusion
  • Repositioning the stent
  • Reimplantation or bypass of the affected vessels
• Work with what perfusion you've got
  • Cease enteral nutrition to reduce the mesenteric blood flow requirements (also, less stool to perforate with)
  • Aim for a higher MAP (~80, or whatever the surgeons are comfortable with)
• Minimise counterperfusionary factors:
  • Minimise dependence on vasopressors that limit splanchnic blood supply
  • Focus on increasing blood pressure through increasing cardiac output (eg. by optimising preload)
  • Decompress the gut with NG drainage, hoping to reduce the transmural pressure
  • Control the abdominal compartment pressure by the conventional methods, or by opening the abdomen
• Percutaneous intervention
  •  Stahl et al (2020) found that local vasodilator therapy appeared technically feasible and potentially beneficial in non-occlusive mesenteric ischaemia, but then this is not usually the type of ischaemia you get from EVAR, so it may not be applicable.
• Remove ischaemic bowel
  • Gut at Grade II or III is likely to have transmural necrosis and probably needs to come out before it makes everything worse
  • Partial or total colectomy is often required
  • The laparotomy gives the option of evacuating any haematoma that has accumulated, or to revascularise any mesenteric vessels that require revascularisation
  • The abdomen can be left open to improve the perfusion of the residual bowel

Alternatively:

• Do nothing ("conservative management")
  • Antibiotics (broad spectrum, anaerobe cover)
  • Bowel rest
  • Analgesia
  • Haemodynamic support
  • Under generic circumstances, Leone et al (2018) found a 14% survival rate from unspecified arterial mesenteric ischaemia, which, the attentive reader will note, is not 0%.