CTA Aortic Stent Graft Protocol: The Complete Guide to Endoleak Surveillance and EVAR CT Imaging

Introduction to post-EVAR CT surveillance

The CTA aortic stent graft protocol is the surveillance backbone of every endovascular aneurysm repair (EVAR) program. Once a fabric-and-metal endograft is deployed to exclude an abdominal or thoracic aortic aneurysm from systemic pressure, the radiology department inherits a lifelong monitoring responsibility: confirming that the sac remains excluded, that the graft limbs remain patent, and that no leak is re-pressurizing the aneurysm wall. A technically inadequate scan does not simply produce a blurry image — it can hide a slow-filling Type II endoleak, miss early graft migration, or falsely suggest endoleak where only post-surgical glue is present.

Unlike a routine contrast CT, the aortic stent graft examination is built around timing rather than enhancement alone. Three acquisitions — unenhanced, arterial, and a fixed 90-second delayed phase — are compared slice-for-slice to determine whether contrast is entering the excluded aneurysm sac. Because the fabric graft and metallic stent skeleton sit directly in the path of the aorta, every phase must also be optimized for metal artifact reduction, since blooming from nitinol struts can both obscure subtle leaks and mimic them. This article walks radiographers, radiologists, and referring physicians through the complete protocol: positioning and phase timing, the full injection parameters, dose benchmarks, the ten pathologies that define EVAR follow-up, and the layered pitfalls that occur at the console, in the reading room, and at the bedside.

Abdominal aortic aneurysm repair has shifted dramatically toward the endovascular approach over the past two decades, and most vascular surgery programs now treat EVAR as the first-line option in anatomically suitable patients. This shift has produced a large and growing population of patients who require structured, often lifelong, imaging follow-up — typically annual or biennial CTA, sometimes supplemented by duplex ultrasound in lower-risk patients once the sac has demonstrated stability. For a busy CT department, this means the aortic stent graft protocol is rarely a one-off scan; it is a recurring relationship with the same patient over years, which makes protocol consistency across visits just as important as protocol accuracy on any single visit. A sac that appears to have grown 4 mm may simply reflect a different slice level or a different contrast bolus geometry between two studies performed on different scanners — which is exactly why this guide emphasizes reproducibility at every step, from patient positioning through to measurement technique at the workstation.

The clinical stakes of getting this protocol right are significant. An undetected high-pressure endoleak can re-pressurize the aneurysm sac years after what was believed to be a successful repair, eventually leading to rupture — a catastrophic, frequently fatal event that the original endovascular procedure was specifically designed to prevent. Conversely, over-calling a benign finding such as residual embolization glue can trigger an unnecessary and not-risk-free re-intervention. Both directions of error are avoidable with disciplined technique, and that is the purpose of this guide: to give every member of the imaging chain — from the technologist who sets up the injector to the physician who reads the final report — a shared, evidence-based understanding of what a correct aortic stent graft CTA looks like, and why each step of the triphasic acquisition exists.

Anatomy and Hounsfield unit reference values

A working knowledge of the device anatomy is just as important as native vascular anatomy for this protocol. Most infrarenal endografts are bifurcated, modular devices consisting of a main body with a suprarenal or infrarenal fixation stent, a contralateral limb gate, and two iliac limbs that are deployed and overlapped to create a continuous conduit. Thoracic devices (TEVAR) are typically single tubular grafts with proximal and distal sealing zones in relatively healthy, non-aneurysmal aorta. The “landing zones” — the segments of native aorta and iliac artery that the graft fabric seals against — are the single most important anatomic concept in this protocol, because the overwhelming majority of Type I endoleaks occur precisely at these attachment sites.

Relevant clinical anatomy

The aneurysm sac is the residual outer wall of the native aorta that remains in situ around the graft after exclusion; it should shrink or stabilize over time if the repair is successful. The perigraft space is the potential space between the fabric and the native wall — this is exactly where endoleak contrast pools and must be distinguished from the true graft lumen. Collateral vessels that classically reconstitute the sac in a Type II leak include the lumbar arteries, the inferior mesenteric artery (IMA), accessory renal arteries, and the median sacral artery. For thoracic grafts, the intercostal and bronchial arteries play the same collateral role. Iliac limb anatomy — particularly tortuosity and the presence of a hypogastric (internal iliac) occluder coil — frequently determines whether a Type Ic or Type II leak from pelvic collaterals develops.

Device types and fixation zones

Standard infrarenal devices rely on a sealing zone of at least 10–15 mm of relatively healthy, non-aneurysmal aorta below the renal arteries; fixation can be suprarenal, with bare uncovered struts extending above the renal ostia for additional anchoring, or infrarenal, where the entire fabric component sits below the renal arteries. This distinction matters directly for image interpretation: with a suprarenal fixation device, the radiologist must trace the bare metal struts as they cross the renal artery origins to confirm renal perfusion has not been compromised, while reporting the sealing zone itself at the first covered fabric segment below. Complex aneurysm anatomy with short or angulated necks is increasingly managed with fenestrated or branched endografts, which incorporate precisely positioned openings or side-branches to preserve flow into the renal, superior mesenteric, and celiac arteries — these devices demand even more meticulous multiplanar review of each fenestration or branch for patency and leak at its individual seal point.

For the radiographer planning coverage, it is essential to know which device type is implanted before the patient is positioned on the table: a fenestrated or branched device requires the field of view to extend cranially enough to capture the visceral segment fenestrations in full, whereas a standard infrarenal tube/bifurcated device does not. Reviewing the operative report or device identification card at check-in — rather than defaulting to a generic abdominal aorta range — is the single fastest way to avoid an incomplete study that has to be repeated.

These reference values are starting points, not absolute cutoffs, and should always be interpreted alongside the patient’s own prior studies. A focal area measuring 55 HU might be entirely normal background sac density in one patient and a clear endoleak in another whose baseline sac has historically measured 25 HU — which is precisely why the non-contrast phase is not an optional step that can be skipped to save dose. It is the individualized reference against which every subsequent measurement on that patient is judged, and it is also the phase that most reliably separates true contrast leak from permanently hyperdense embolization material, since glue and coils are already dense before any contrast is given.

Scanning technique: 7 steps to a diagnostic study

1. Patient preparation and positioning. Supine, feet first or head first per departmental convention, arms raised above the head to remove beam-hardening streak artifact across the abdomen and pelvis. Confirm prior operative report or device card to identify graft type, landing zones, and any known prior endoleak. For patients with limited shoulder mobility, a single-arm-raised position with the contralateral arm crossed over the chest is an acceptable compromise that still substantially reduces streak artifact compared with both arms down.
2. Scout and coverage planning. Acquire a full anteroposterior scout from the diaphragm to the lesser trochanters to ensure both iliac limbs and the distal landing zones are included; thoracic grafts require coverage from the thoracic inlet to the celiac axis or below, depending on graft length. Cross-check the planned range against the device card or operative note rather than relying on a generic abdominal aorta preset, since fenestrated and branched devices frequently extend further cranially than a standard infrarenal graft.
3. Unenhanced (non-contrast) acquisition. A low-dose non-contrast series establishes the sac baseline density, identifies calcification and prior embolization material, and is mandatory for distinguishing hyperdense glue from a true leak on the delayed phase. Reduced tube current is appropriate here, since this phase is used primarily for density reference and bone/calcium mapping rather than fine soft-tissue contrast resolution.
4. Bolus-tracking placement and arterial phase. Position the region of interest in the supraceliac or proximal abdominal aorta; trigger threshold and table movement are detailed in the contrast section below. Reconstruct at <1.0 mm slice thickness for multiplanar and 3D reformatting of the landing zones and limb overlap segments. Coach the patient through a single, reproducible breath-hold instruction immediately before the trigger fires, since inconsistent breath-holding between the arterial and delayed phases is a common source of slice-misregistration when comparing the two series side by side.
5. Fixed delayed-phase acquisition at 90 seconds. This is the single most important step for endoleak detection — slow-filling Type II leaks frequently are isodense or invisible on the arterial phase and only declare themselves once equilibrium contrast has had time to pool in the perigraft space. Use a programmed timer rather than visual estimation, and keep the patient on the table in the same position used for the arterial phase to minimize repositioning artifact between the two enhanced series.
6. Image reconstruction. Generate thin-slice axial, coronal, and sagittal multiplanar reformats plus a maximum-intensity-projection (MIP) and 3D volume-rendered reconstruction of the graft for surgical planning and limb patency assessment; apply a metal artifact reduction (MAR) algorithm to suppress strut blooming. Where dual-energy or photon-counting hardware is available, generate the iodine map and/or virtual non-contrast series alongside the conventional reconstructions at this step rather than as a separate, easily forgotten add-on later.
7. Quality control and comparison. Cross-reference sac diameter and HU values against the prior surveillance study at the workstation before the patient leaves the department, since a repeat delayed phase is far easier to add immediately than to recall the patient. If a measurement appears discordant with the prior study by more than a few millimeters, re-examine whether the comparison was made at the same anatomic level before assuming true interval growth.

Scanner generation comparison (16-slice to 320-slice)

The practical takeaway across all scanner generations is that z-axis coverage speed, not raw slice count, is what most directly affects image quality for this protocol. A patient with limited breath-hold capacity scanned on an older 16-slice system is far more likely to produce a misregistered, motion-degraded comparison between the arterial and delayed phases than the same patient scanned on a wide-detector system capable of covering the full aorto-iliac segment in a single rotation. Departments running mixed fleets should document expected coverage times per scanner class and adjust breath-hold coaching accordingly rather than using one script for every machine.

Dual-energy and photon-counting protocol considerations

Deep learning reconstruction (DLR)

Deep learning reconstruction algorithms are increasingly used to denoise low-dose aortic surveillance acquisitions, allowing departments to reduce tube current on the non-contrast and delayed phases — the two phases that contribute the most cumulative dose with the least diagnostic novelty — while preserving the sharp edge detail needed to measure sac diameter precisely. Vendors report that DLR can support meaningful mA reduction on follow-up scans without compromising endoleak detection sensitivity, though departments should validate this locally against their reference reading workflow before lowering dose protocols.

For a lifelong surveillance population, the cumulative dose argument for DLR is particularly compelling: a patient undergoing annual triphasic CT for ten or more years following EVAR accumulates a substantial lifetime effective dose, and any per-study reduction compounds meaningfully over that timeframe. Departments adopting DLR for this protocol typically phase in mA reduction gradually, comparing a subset of DLR-reconstructed low-dose studies against standard-dose comparators on the same patients before fully transitioning the protocol, to confirm that small Type II endoleaks and subtle sac diameter changes remain reliably detectable at the reduced exposure.

Contrast media injection protocol

Because this protocol depends on precise timing across three phases, contrast delivery consistency is as important as scanner technology. A 90 mL bolus of iodinated contrast is delivered at 4.0 mL/s through an 18–20 gauge antecubital cannula, immediately followed by a 100 mL saline chaser at a matched flow rate to clear the injection line of residual contrast and maintain a tight, compact bolus.

The 4.0 mL/s flow rate sits at the higher end of typical departmental ranges and is selected specifically to produce a sharply defined arterial bolus despite the relatively modest 90 mL total volume — a deliberate trade-off that keeps total iodine load lower across a lifetime of repeated annual studies while still achieving diagnostic aortic opacification above 250 HU. The saline chaser is not a minor afterthought: at this flow rate, a substantial fraction of the contrast bolus would otherwise remain in the injection tubing and peripheral veins rather than reaching the central circulation, blunting peak aortic enhancement and undermining bolus-tracking trigger accuracy.

Bolus tracking should use a trigger threshold appropriate to the local injector and scanner combination — most departments set this in the supraceliac or proximal abdominal aorta at a threshold in the 100–150 HU range above baseline, with a short diagnostic delay (typically 4–8 seconds) built in after the trigger fires to allow the table to reach the scan start position and the patient to reach full inspiration. Because EVAR patients are frequently elderly with reduced cardiac output, bolus arrival can be slower and more variable than in a younger trauma or oncology population; technologists should be prepared to extend the monitoring window slightly if the trigger threshold has not been reached within the expected timeframe rather than assuming injector or line failure.

Patient screening and premedication

Standard iodinated contrast screening applies in full: a documented history of prior contrast reaction, asthma, or significant atopy should prompt premedication per departmental protocol or selection of an alternative imaging strategy where contrast cannot be safely given. Because this patient population is typically older with a higher burden of cardiovascular and renal comorbidity than a general CT referral population, renal function screening deserves particular emphasis — see the safety callout below.

Extravasation management

High-flow power injection carries an inherent extravasation risk, and EVAR patients with fragile or previously cannulated veins from years of prior surveillance studies may be at elevated risk compared with a first-time contrast patient. A pre-injection test flush with 8–10 mL of saline at the planned flow rate, combined with direct visual and palpable monitoring of the injection site for the first several seconds of the bolus, remains the most effective bedside safeguard. Should extravasation occur, the department’s standard protocol — typically limb elevation, ice application, and clinical monitoring with escalation to specialist referral for large-volume extravasations — should be followed, and the incomplete study should be rescheduled rather than interpreted from a compromised bolus.

Radiation dose and dose reference levels

Triphasic acquisitions are inherently dose-intensive, so EVAR surveillance protocols are a frequent target for optimization, particularly because many patients undergo this scan annually for years. Diagnostic reference levels (DRLs) should be benchmarked locally and reviewed against national registries, in alignment with European Commission RP 185, AAPM dose index reporting guidance, and ICRP optimization principles.

It is worth contextualizing these figures for patients and referring clinicians who may ask about cumulative exposure: a single triphasic surveillance study delivers an effective dose roughly comparable to several years of average background environmental radiation, and a patient undergoing ten consecutive annual studies will accumulate a non-trivial lifetime dose. This is precisely why surveillance interval and protocol optimization — not just per-scan technique — are part of responsible long-term EVAR management, and why many vascular programs now shift stable, leak-free patients to less frequent CT with interval duplex ultrasound once a low-risk trajectory has been established.

SSDE calculation deserves particular attention in this population because body habitus varies enormously among EVAR patients, many of whom carry significant cardiovascular comorbidity associated with both larger and smaller body sizes. A standard-sized CTDIvol value can substantially under- or overestimate the dose actually absorbed by a given patient; converting to SSDE using the patient’s measured lateral or AP diameter at the workstation, as recommended by AAPM Report 204, gives a far more clinically meaningful figure for dose tracking and for comparison against departmental DRLs over time.

5 dose reduction strategies

• Replace true non-contrast with virtual non-contrast (VNC) using dual-energy acquisition where validated locally, removing one full phase of exposure.
• Apply automatic tube current modulation referenced to patient size on every phase, not just the arterial acquisition.
• Adopt deep learning or iterative reconstruction to permit lower mA on the non-contrast and delayed phases.
• Tighten z-axis coverage to the known graft length plus a margin, rather than scanning the full abdomen and pelvis by default on every follow-up.
• Risk-stratify surveillance intervals so that patients with a stable, leak-free sac on two consecutive studies move to ultrasound-based or extended-interval CT follow-up rather than annual full triphasic CT indefinitely.

Top 10 pathologies on aortic stent graft CTA

From a management standpoint, these ten findings naturally separate into three tiers of urgency that should shape both the radiology report and the referral pathway. High-pressure leaks — Type I and Type III — carry a direct rupture risk and are generally flagged for prompt vascular surgery or interventional radiology review regardless of sac size. Pressure-neutral or self-limiting findings — most Type II leaks and Type IV leaks — are typically managed with continued surveillance, with intervention reserved for cases demonstrating associated sac growth. Structural and infectious complications — graft migration, infection, thrombosis, and limb occlusion — each carry their own distinct management algorithm independent of leak status, and should never be subsumed under a generic “endoleak” label in either the report or the referral conversation. Sac expansion, finally, functions as the universal red flag that elevates any of the above findings to urgent regardless of how reassuring the rest of the study appears.

Pitfalls for radiographers

The primary scanning pitfall in this protocol is omitting the delayed imaging phase. Because Type II endoleaks fill slowly through small collateral vessels, relying strictly on the arterial phase — the phase most technologists are conditioned to prioritize on every other CTA protocol — will completely miss leaks that only become visible once equilibrium contrast has had time to pool in the perigraft space at 90 seconds or later.

This pitfall is particularly insidious because the resulting images look entirely normal: the aorta and graft limbs are well opacified, the arterial-phase reconstructions are technically excellent, and nothing in the arterial series itself signals that anything has been missed. The absence is only apparent in retrospect, once a delayed-phase comparison from a subsequent study reveals a leak that, in hindsight, was almost certainly present — and almost certainly missable — on the prior incomplete study. Building the delayed phase into the protocol as a hard-coded, non-skippable step removes the dependence on individual technologist vigilance for a finding pattern that, by its nature, gives no warning sign at the console.

Pitfalls for radiologists

The primary interpretation pitfall in this protocol is misidentifying residual high-density surgical glue or calcification within the aneurysm wall as an active endoleak. Many EVAR patients have undergone prior adjunctive embolization with N-butyl cyanoacrylate glue or coils, which remain permanently hyperdense on every subsequent surveillance scan and can closely mimic the appearance of contrast pooling if the radiologist is not actively comparing density across phases.

The distinguishing feature is straightforward once it is actively sought: true endoleak material is contrast, and contrast by definition is only present after intravenous injection — meaning it should be absent, or at most reflect background sac density, on the non-contrast baseline phase. Embolization glue and coils, by contrast, are already radiopaque before any contrast is given and remain essentially unchanged in density across all three phases. A radiologist who scrolls directly to the delayed phase without first reviewing the non-contrast series loses this single most reliable discriminator, which is why a disciplined, phase-by-phase reading order — rather than a delayed-phase-only quick read — remains essential for this protocol.

Pitfalls for non-radiology physicians

Referring clinicians — primary care physicians, emergency department staff, and non-vascular specialists who encounter EVAR patients incidentally — are not expected to interpret CTA images, but they are frequently the first point of contact when a surveillance report lands in the inbox or when a patient presents with new symptoms. The pitfalls below center on translation gaps between a technically precise radiology report and the clinical action it should trigger.

Pitfall comparison summary

AI and automation in EVAR surveillance

Automated sac volumetry and centerline measurement tools are increasingly used to remove the inter-observer variability that plagues manual axial diameter measurement, directly addressing one of the most consequential pitfalls in this protocol. Several CE-marked and FDA-cleared platforms now offer semi-automated 3D sac segmentation, automatic comparison to prior studies at matched anatomic levels, and quantitative growth-trend reporting that flags sac expansion before it would be apparent on a single side-by-side visual read. Evidence in the vascular literature supports volumetric sac measurement as more sensitive to early growth than simple maximal-diameter calipering, particularly in irregularly shaped or partially thrombosed sacs.

Beyond volumetry, automated centerline-based limb patency analysis can flag focal luminal narrowing or occlusion along a tortuous iliac limb that might otherwise require painstaking manual scrolling through hundreds of axial images, and some platforms now incorporate iodine-map overlays from dual-energy acquisitions directly into the segmentation output, helping the reader visually separate low-attenuation collateral filling from background sac density at a glance. Structured reporting templates that auto-populate endoleak subtype, sac volume trend, and limb patency status into a standardized macro also reduce the risk that a referring physician receives an ambiguous free-text report — directly mitigating the clinical-pitfall pattern described in the section above, where “endoleak” is treated as a single undifferentiated diagnosis rather than a specific, actionable subtype.

Typical surveillance intervals by finding

While exact intervals vary by institutional protocol and should always follow the treating vascular surgeon’s recommendation, the table below summarizes commonly used surveillance patterns that referring teams can use as a general orientation when reading a report.

Further reading

The following SATMED Health resources cover adjacent CTA technique, contrast delivery safety, and multi-phase protocol design topics that complement the aortic stent graft surveillance workflow described above.

1. CTA Lower Extremity Runoff: 7 Critical Protocol Steps — bolus-tracking parameters and calcium blooming pitfalls relevant to any peripheral or aortic CTA.
2. 5 Master Mesenteric CTA Protocol Tactics — dual-phase arterial timing and reconstruction strategy applicable to aortic branch vessel evaluation.
3. 7 Critical CTA Brain & Carotids Protocol Steps Every Radiographer Must Master — bolus-tracking and venous contamination pitfalls common to all time-critical CTA studies.
4. 7 Critical Adrenal Washout CT Protocol Steps — a comparable multi-phase, delayed-imaging protocol structure for benchmarking timing discipline.
5. The Price We Pay for Bubbles in CT and MRI: Understanding Venous Air Embolism in Contrast-Enhanced Imaging — essential safety reading for any high-flow power injection protocol, including this one.

Conclusion

The CTA aortic stent graft protocol succeeds or fails on phase discipline. A technically complete triphasic acquisition — unenhanced baseline, bolus-tracked arterial phase, and a fixed 90-second delayed phase — is what allows the department to confidently separate the ten core findings covered in this guide, from the high-urgency Type I and Type III endoleaks that demand same-week surgical referral, to the often-observed Type II leak, to graft migration, infection, thrombosis, sac expansion, and limb occlusion. The two pitfalls that recur most often in practice sit at opposite ends of the workflow: the radiographer who stops scanning before the delayed phase, and the radiologist who mistakes old embolization glue for a new leak. Closing both gaps — through protocol-card enforcement at the console and disciplined phase-by-phase comparison at the workstation — is what keeps lifelong EVAR surveillance both diagnostically reliable and radiation-conscious.

Because this is a surveillance protocol performed repeatedly on the same patient over many years, the cumulative value of getting every step right compounds with each visit. A department that standardizes patient positioning, bolus-tracking thresholds, delayed-phase timing, and measurement technique builds a dataset of genuinely comparable studies — one where a 4 mm change in sac diameter reflects true biology rather than technique drift between visits. That consistency, paired with disciplined reporting that names the specific endoleak subtype and trend rather than a generic “endoleak present,” is ultimately what protects EVAR patients from both the under-recognized rupture risk of a missed leak and the unnecessary intervention risk of a misread one.

References