Cerebral Embolization, Silent Cerebral Infarction and Neurocognitive Decline After

Thoracic Endovascular Aortic Repair

Anisha H Perera MRCS1, Nung Rudarakanchana PhD FRCS2, Leonardo Monzon FRCR3,

Colin D Bicknell MD FRCS1, Bijan Modarai PhD FRCS4, O Kirmi FRCR5, Thanos

Athanasiou MD PhD MBA FECTS FRCS6, Mohammad Hamady MD FRCR1, 7, Richard G

Gibbs MD FRCS1

1 Imperial Vascular Unit, Department of Surgery and Cancer, Imperial College and Imperial

Healthcare NHS Trust, London, UK

2 Department of Vascular Surgery, Royal Free Hospital, London, UK

3 Department of Interventional Radiology, Guy’s and St Thomas’ NHS Foundation Trust,

London, UK

4 Academic Department of Vascular Surgery, King’s College London, BHF Centre of

Research Excellence & NIHR Biomedical Research Centre at King’s Health Partners, St

Thomas’ Hospital, London, UK

5 Department of Neuroradiology, Imperial Healthcare NHS Trust, London, UK

6 Department of Surgery, Imperial College London, London, UK

7 Department of Interventional Radiology, Imperial Healthcare NHS Trust, London, UK

Corresponding author Colin Bicknell

Department of Surgery, Imperial College London, 10th floor QEQM, St Mary’s Hospital,

South Wharf Road, London, W2 1NY, UK

Telephone: +44 (0) 20 3312 6666, Email: colin.bicknell@imperial.ac.uk

1

Sources of funding This study was supported by the National Institute for Health Research

(NIHR) Biomedical Research Centre based at Imperial College Healthcare NHS Trust and

Imperial College London.

Category Original article

Previous communication Silent cerebral infarction and neurocognitive decline following

thoracic endovascular aortic repair. Oral presentation Vascular Society AGM scientific prize

session, Glasgow November 2014. Br J Surg. 2015; 102 (suppl. 2): 5.

2

ABSTRACT

Background

Silent cerebral infarction is brain injury detected incidentally on imaging, which can be

associated with cognitive decline and future stroke. This study investigates cerebral

embolization, silent cerebral infarction and neurocognitive decline following thoracic

endovascular aortic repair (TEVAR).

Method

Patients undergoing elective or emergency TEVAR at Imperial College Healthcare NHS

Trust and Guys’ and St Thomas’ Hospital NHS Trust between January 2012 and April 2015

were recruited. Aortic atheroma graded 1 (normal) to 5 (mobile atheroma) was evaluated on

pre-operative computed tomography. Patients underwent intra-operative transcranial Doppler

(TCD), pre- and post-operative cerebral magnetic resonance imaging (MRI) and

neurocognitive assessment.

Results

Fifty-two patients underwent TEVAR. Higher rates of TCD-detected embolization were

detected with greater aortic atheroma (grade 4-5 median 207 versus grade 1-3 median 100,

p=0.042), more proximal landing zones (zone 0-1=median 450 versus zones 3-4=median 72,

p=0.001) and during stent-graft deployment and contrast injection (p=0.001). On univariable

analysis, left subclavian bypass (ß Coefficient=0.423, standard error=132.62, p=0.005),

proximal landing zone 0-1 (ßcoef=0.504, SE=170.57, p=0.001) and arch hybrid procedure

(ßcoef=0.514, SE=182.96, p=0.000) were predictors of cerebral emboli. Cerebral infarction

was detected in 25/31(81%) undergoing MRI: 21 silent (68%) and four clinical strokes

3

(13%). Neurocognitive decline was seen in 6/7 domains assessed in 15 patients with silent

cerebral infarction, with age a significant predictor of decline.

Conclusion

This study demonstrates a high rate of cerebral embolization and neurocognitive decline

affecting patients following TEVAR. Brain injury following TEVAR is more common than

previously recognised, with silent or overt stroke in over 80% of patients.

4

MANUSCRIPT

INTRODUCTION

Thoracic endovascular aortic repair (TEVAR) has improved survival outcomes in patients

with thoracic aortic disease (TAD) over open surgery, with reported elective 30-day mortality

rates of 3-8%1-4. Neurological complications of stroke and paraplegia remain a significant risk

however. Stroke occurs in up to 7% of patients following TEVAR and is associated with a

significant increase in post-operative morbidity and mortality1, 4-6. Strokes are mainly

embolic4, 5 as endovascular repair involves the advancement of stiff wires, catheters and

delivery devices (up to 26Fr in size) through the aorta often diseased with significant

atheroma. Risk factors for stroke include previous stroke, a more proximal extent of repair,

high-grade aortic arch atheroma, intra-operative hypotension, chronic renal insufficiency and

coverage of the left subclavian artery (LSCA) without revascularization (4-8). Only two

studies to date have investigated cerebral embolization during TEVAR using transcranial

Doppler (TCD)9, 10.

Peri-operative embolization may also cause silent cerebral infarction. This is imaging

evidence of cerebral infarction without acute neurologic dysfunction attributable to the

lesion11, and is commonly investigated using magnetic resonance imaging (MRI). Studies

have shown it is associated with an increased risk of future stroke, dementia, depression,

cognitive impairment and early mortality independent of other vascular risk factors12, 13. It is

well recognized following other catheter-based procedures (cerebral14,15 and coronary16

angiography, carotid stenting17 and transcatheter aortic valve implantation18), and a recent

study of 19 patients reported a 63% incidence following TEVAR19.

5

This study was performed to establish the incidence of TCD-detected cerebral embolization,

silent cerebral infarction and neurocognitive decline following TEVAR, and evaluate patient

and procedural risk factors for cerebral embolization and silent infarction.

6

METHODS

Patient population

Ethical approval was obtained from the UK National Research Ethics Service Committee

London- Fulham (08/H0711/59) and all patients gave written informed consent. Patients

undergoing TEVAR as elective or emergency for any thoracic aortic pathology at two tertiary

referral vascular surgery centres (St Mary’s Hospital, Imperial College Healthcare National

Health Service Trust and St Thomas’ Hospital, Guys’ and St Thomas’ Hospital National

Health Service Foundation Trust) between January 2012 and April 2015 were eligible for

inclusion.

Pre-operative evaluation of aortic atheroma

All patients underwent computed tomography angiography (CTA) prior to surgery in

accordance with standard clinical protocol with 1mm slice thickness and intravenous

iodinated contrast. Multi-planar 3-dimensional image reconstruction was performed on

dedicated CT workstations (Extended Brilliance Workspace, V3.5.0.2254, Philips Medical

Systems). Two independent interventional radiologists conducted evaluation of the arch and

descending thoracic aorta for the presence of aortic atheroma, calcification and mural

thrombus. Atheroma was quantitatively graded based on thickness: grade 1= normal, grade

2= intimal thickening, grade 3= atheroma </=5 mm, grade 4= atheroma >5 mm, and grade 5=

mobile lesion7, 20. Aortic arch type based on the relationship of the origins of the supra-aortic

vessels to the parallel plane perpendicular to the outer curvature of the arch (I, II, III or

bovine)8 and proximal landing zone (PLZ) according to Ishimaru’s classification (zone 0-4)

was also assessed21. All elective patients underwent carotid and vertebral artery duplex by

qualified Vascular Sonographers.The circle of willis was not assessed on CTA.

7

Cerebral MRI

Cerebral MRI was performed using the Discovery MR750w system (GE Healthcare, United

Kingdom) with 3.0T magnet strength; axial Diffusion Weighted Imaging (DWI) and axial

T2-weighted fluid-attenuated inversion recovery (FLAIR) with 5mm slice thickness. MRI

was performed pre-operatively to identify any pre-existing infarcts, and repeated post-

operatively within 72 hours or as soon as clinical condition of the patient allowed. Sensitivity

and specificity of DWI in the detection of acute cerebral ischemia is 99% and 92%

respectively22. A bright DWI signal indicates an early or sub-acute, but not chronic (>14 days

old) ischaemic lesion. Two independent neuroradiologists blinded to clinical outcomes

compared pre- and post-operative MRIs, reporting the number, laterality and vascular

territory (anterior or posterior circulation or border-zone territory) of new lesions. Lesion

surface area was measured on the slice on which the lesion had the largest diameter.

Neurocognitive testing

Neurocognitive testing was performed by a trained assessor pre-operatively (one day prior to

TEVAR), post-operatively (when medically fit for discharge), and at first follow-up

outpatient appointment (OPA), approximately six to eight weeks post-discharge. The test

battery included the following tests which are the recommended core neuropsychological

battery for assessment of neurobehavioral outcomes after cardiac surgery23: 1) REY auditory

verbal learning test (verbal learning and memory), 2) Trail making test A (visual search and

motor skills), 3) Trail making test B (higher level cognitive skills e.g. mental flexibility and

switching), and 4) grooved pegboard (manual dexterity, fine motor skills and complex visual-

motor co-ordination of both the dominant and non-dominant hand). The controlled oral word

association test (COWA, FAS version), a measure of executive cognitive dysfunction, was

also included.

8

TEVAR (intervention group)

All cases were performed under general anaesthesia (GA) at both centers, with intravenous

heparin maintaining an activated clotting time of 250-350 seconds. In patients where

satisfactory proximal sealing required coverage of the LSCA, the LSCA was maintained by

left carotid-subclavian bypass or a proximal scalloped endograft custom-made in

anatomically suitable cases. Extra-anatomical bypass with visceral debranching (visceral

hybrid) or TEVAR with left carotid-subclavian bypass was performed as a single-stage

procedure, whilst full arch debranching (left carotid-subclavian bypass and carotid-carotid

bypass: arch hybrid) was staged. A spinal drainage catheter was inserted pre-operatively in all

patients undergoing visceral hybrid or coverage of >25cm or >50% of aortic length. Access

was via percutaneous common femoral artery (CFA) puncture or open surgical groin cut

down, and a 22Fr sheath was inserted. Contralateral percutaneous CFA access was obtained

with a 5Fr sheath for insertion of an imaging catheter. Pharmacological blood pressure

manipulation using glyceryl trinitrate infusion was employed to achieve controlled systemic

hypotension with a systolic blood pressure of 80mmHg for a short period of time only to aid

precise deployment. All procedures were performed in a dedicated vascular hybrid suite by a

team which included both an experienced vascular surgeon and interntional radiology

consultant.

Control group

Patients undergoing endovascular aortic repair (EVAR) or fenestrated endovascular aortic

repair (FEVAR) of infra-renal or juxtra-renal abdominal aortic aneurysms (AAA) were

recruited as a contemporaneous control group. All cases were performed under GA, with

intravenous heparin maintaining an activated clotting time of 250-350 seconds. Access was

9

via bilateral open surgical femoral access. In all control patients wires and catheters were

placed in the proximal descending thoracic aorta and did not cross the aortic arch.

Intra-operative TCD assessment

Cerebral embolization detection was performed using bilateral TCD insonation of the middle

cerebral artery (MCA). The TCD parameters included DWL® Doppler device with 2-channel

emboli detection software and 2MHz transducer probes, sample volume 8mm, median

softgain 25 (range 13-38), filter setting150Hz and intensity detection threshold >9dB. TCD

signal was continuously recorded during procedures and manual offline analysis performed

by a trained observer for identification of high intensity transient signals (HITS)

characteristic of an embolus. Accepted criteria for emboli detection was used: alteration of

Doppler signal to a high intensity (>3dB higher than background blood flow signal), short

duration (<300 milliseconds) and unidirectional signals in the direction of flow accompanied

by a characteristic clicking sound24, 25 (figure 1). Differentiation between gas and solid emboli

was not conducted. A period of pre-operative monitoring was performed after induction of

GA and before commencement of TEVAR to detect any spontaneous embolization. TCD was

performed during carotid-subclavian bypass but not during full arch debranching for arch

hybrids.

Statistical analysis

Statistical analysis was performed using SPSS software (version 22; SPSS, Chicago, IL).

Continuous variables are presented as median (interquartile range, IQR) and categoric

variables as frequency (percentage). Comparisons were made using Mann-Whitney U,

Kruskal-Wallis, Wilcoxon Signed-Rank and Friedman’s test with p<0.05 considered

statistically significant. Logarithmic transformation of non-normally distributed variables was

10

used where required. Univariable and multivariable analysis was performed with entry

criteria into multivariate analysis p<0.05. Univariable analysis for continuous variables was

performed as a linear regression analysis and for binary variables as logistic regression.

RESULTS

Some fifty-two patients undergoing TEVAR and 24 control patients were recruited (table 1).

Clinical outcomes

All stent-grafts were deployed satisfactorily, with no retrograde type A dissection or open

surgical conversion. No endoleak was identified on completion angiography. Median hospital

stay was 8.5 days (interquartile range 4-14), and median follow-up duration 16 months (IQR

6-27). Morbidity and mortality outcomes are detailed in table 2. Four patients (all male, two

elderly) developed clinical stroke (7.7%), all detected immediately post-operatively (table 3).

Procedures performed included one arch hybrid, two TEVAR with carotid-LSCA bypass, and

one proximal stent extension for type Ia endoleak following previous visceral hybrid repair of

type II thoracoabdominal aortic aneurysm. Two patients made a complete recovery following

a period of rehabilitation. No patient with stroke suffered in-patient or 30-day mortality.

Intra-operative TCD-detected cerebral embolization

TCD monitoring was performed in 42 patients; bilateral insonation in 30 patients, with left

only in six patients and right only in six patients due to inadequate bilateral acoustical

temporal bone window. The MCA was insonated in all patients (median depth 50mm, range

46-58mm) except for four where no MCA signal was identified on the right and therefore the

posterior cerebral artery was insonated instead. Cerebral embolization was detected during

TEVAR in all 42 patients. A 53-year old male undergoing emergency TEVAR and left

11

carotid-subclavian bypass for acute type B dissection was the only patient in whom

spontaneous embolization was detected during pre-operative monitoring, where two emboli

were detected at the left MCA. The median procedural embolization rate for all patients was

179 (IQR 71-262), and excluding contrast injection related embolization was 104 (IQR 35-

192). Embolization resulting from contrast injection was included in all analyses. Stent-graft

manipulation and deployment (median 62, IQR 22-163) followed by contrast injection

(median 61, IQR 18-117) resulted in the highest rates of embolization (p=0.001) (figure 2).

Embolization varied depending on patient and procedural factors (figure 3), and was detected

more frequently at the left MCA (p=0.018), with higher aortic atheroma grades 4 and 5

(p=0.042), at more proximal landing zones 0 and 1 (p=0.001), and during arch hybrid

procedures (p=0.001). No significant difference in embolization rates were seen between the

pathologies treated. In the control group ten patients had TCD assessment. In eight EVARs

TCD HITS were detected in five patients; median 1, IQR 0-1.5, range 0-2, with 6 and 16

HITS detected during two FEVARs.

Cerebral DW-MRI

Thirty-one patients underwent pre- and post-operative cerebral MRI (table 3). Not all patients

completed the pre- and post-operative MRI protocol for reasons including emergency

presentation with unstable patient (n=6), significant post-operative morbidity or death (n=5),

patients declining MRI due to claustrophobia (n=7) and contra-indication due to pacemaker

(n=3). Post-operative scans were performed at median four days (IQR 2-6) following

TEVAR. Twenty-five patients (81%) had evidence of brain injury: 21 (68%) had procedure-

related silent infarction (figure 4) and four patients had clinical stroke (13%). All strokes

were confirmed as embolic on MRI or CT. There was a trend towards higher rates of

embolization in patients with silent infarcts compared to those with no post-operative lesions,

12

with the highest rates of embolization in those with clinical stroke (figure 3b). In 21 patients

with silent infarction, there was a total of 120 new lesions detected (median 2 per patient,

IQR 1-9) with a total silent infarct surface area of 5459mm2 (median 16mm2, IQR 9-81.5).

The majority of lesions were left-sided and in the posterior or border-zone circulation. In the

control group seven patients (six EVAR and one FEVAR) underwent pre- and post-operative

MRI, and no patient had new silent infarction or stroke.

Risk factors for cerebral embolization and brain injury

Risk factors identified as predictors of TCD-detected cerebral embolization are detailed in

table 4. The development of brain injury, both silent and stroke (p=0.510), and MRI lesion

surface area (p=0.917) were not shown to be associated with embolization rate. No factors

were independent predictors of embolization on multivariable analysis. Similar results were

seen with TCD HITS excluding contrast embolization, with the same factors associated with

higher embolization rates. On univariable analysis, smoking (odds ratio 8.00, 95% confidence

interval 1.13-56.80, p=0.038) was the only factor significantly associated with the

development of brain injury (both silent and stroke) as determined by new MRI lesions.

Neurocognitive outcomes

Neurocognitive testing was performed in 17 patients (figure 5). Post-operative assessment

testing was conducted at a median time of eight days (IQR 5-13) post-TEVAR. Not all

patients completed neurocognitive testing due to a change in the test battery mid-way during

the study (n=24), emergency presentation with unstable patient (n=6) and significant post-

operative morbidity or death (n=5). In 15 patients with silent cerebral infarction, significant

post-operative neurocognitive decline was seen in 6/7 domains assessed (all except Trails B

test). Univariable analysis revealed age to be a significant predictor of decline in all seven

13

domains, male sex in five domains and TCD HITS (p=0.009) with memory and recall (table

5, supplementary data). To assess the effect of silent cerebral infarction on neurocognition,

patients were divided into three groups; two patients who did not develop silent infarcts, eight

patients with silent infarcts aged<69 years, and seven patients with silent infarcts aged=/>70

years. The median age of patients undergoing neurocognitive testing was 69 years, and this

was chosen as the cut-off value between the two age groups. Patients in the older age group

showed significant post-operative decline in all areas of neurocognition except Trails B test.

This deficit persisted at follow-up OPA for the COWA test (p=0.050) and Trails A (p=0.028),

with median follow-up results remaining below median pre-operative levels for both

components of the auditory verbal learning test. Patients in the younger age group showed

measurable and significant decline affecting both the dominant (p=0.018) and non-dominant

hand (p=0.039) in the manual dexterity pegboard test, with persistent decline present at

follow-up affecting the dominant hand (p=0.028). In the control group seven patients (5

EVAR and 2 FEVAR) underwent neurocognitive testing at the three time-points.

Neurocognitive decline was seen in only 1/7 domains tested; COWA (p=0.002). The decline

was seen between pre- and post-operative testing (p=0.018) and between pre-operative and

OPA testing (p=0.027).

14

DISCUSSION

These results demonstrate TCD-detected cerebral embolization in all patients during TEVAR,

with an overall brain injury (both silent and stoke) rate of 81% in patients undergoing MRI.

The overall stroke rate was 7.7%, as this cohort includes high-risk patients requiring more

proximal landing zones, with several undergoing arch hybrids and LSCA revascularization.

The silent infarction rate of 68% is comparable to the 63% rate detected by Kahlert and

colleagues19.

Greater aortic atheroma burden, LSCA bypass, more proximal landing zones 0 and 1, and

arch hybrid repairs were shown to be predictors of increased embolization. This is likely due

to the more proximal landing zones which is a documented risk factor for stroke following

TEVAR, increased instrumentation of arch vessels and higher atheroma load and mobile

lesions contributing to increased embolization. There were minimal HITS and no silent

infarcts in patients undergoing EVAR and FEVAR, likely due to the PLZ in the abdominal

aorta and limited instrumentation in the aortic arch in the control patients. There was a trend

towards a higher number of TCD HITS in patients who developed stroke, followed by

patients with silent infarcts, with the lowest HITS in those with no post-operative lesions.

There was no association however between number of TCD HITS and silent cerebral

infarction. This is contrary to the findings of Bismuth and colleagues who showed a

15

significant association between total HITS and post-operative stroke9. In that study of 20

patients undergoing TCD during TEVAR, the highest number of HITS was generated by

catheter manipulation during the diagnostic phase and device deployment during the

treatment phase. HITS during contrast injection were not examined. This differs from the

present findings, where contrast injection contributed the highest HITS during the diagnostic

phase and stent-graft deployment the highest HITS during the treatment phase. In a study

investigating cerebral embolization during cerebral angiography, 150 patients were

randomized into three groups: control group with conventional angiography, heparin group

with systemic heparinization throughout the procedure, and an air filter group with an air

filter between the catheter and contrast syringe26. There were equal numbers of MRI lesions

in the heparin and air filter groups, which were significantly fewer than in the control group

(p=0.002). TCD HITS were significantly lower in the air filter group (p<0.01) compared to

the heparin and control groups, which did not differ from each other. This demonstrated the

role of air embolization related to contrast injection and catheter flushing in the development

of silent cerebral infarcts, and highlights the importance of reducing both gaseous and solid

embolization. The results presented are in keeping with this, and show a high number of TCD

HITS generated during both contrast injection and flushing. In addition, analysis of risk

factors for embolization revealed comparable results both including and excluding contrast

injection, and therefore results should include and be interpreted with embolization during

contrast injection.

Routine clinical examination assesses for focal neurological abnormalities including paresis,

ataxia, visual defects and aphasia. Global brain dysfunction such as cognitive decline,

memory and mood disturbances, reduction of psychomotor speed and personality changes

may be missed because they require specific neurocognitive tests for diagnosis15, 27. In 2,229

16

Framingham Offspring Study participants silent cerebral infarcts predicted an increased risk

of stroke and dementia independent of vascular risk factors, and also indicated an increased

risk of mild cognitive impairment and death12. People with silent infarcts have more than

double the risk of dementia, and in particular Alzheimer’s disease, compared to the general

population28. Post-operative cognitive dysfunction (POCD) after GA and surgery is a

recognised phenomenon with multifactorial aetiology29. In patients undergoing open cardiac

surgery cognitive decline was seen in all patients with post-operative DW-MRI ischemic

lesions, and 35% without imaging-identified lesions (p<0.001)30. There was also an

association between the number of abnormal cognitive tests and ischemic burden (p<0.001)30.

This demonstrates cognitive impairment is associated with peri-operative ischemia and

degree of ischemic load, and not simply a consequence of POCD. One TAVI study showed

that age was independently associated with longitudinal cognitive decline31. The

neurocognitive control group consisted of an elderly cohort with a decline in only 1/7 areas,

whereas the similarly elderly TEVAR group had a decline in 6/7 domains. This highlights

that while elderly patients may be affected by GA-related POCD, the significant decline seen

in multiple domains post-TEVAR is most likely a consequence of cerebral ischemic burden.

Few studies have investigated neurocognitive decline following endovascular procedures to

date. Two studies investigating silent cerebral infarction following coronary catheterization

found a significant decline in learning and attention and verbal and non-verbal memory

associated with silent infarcts32, 33. In this study follow-up assessment was performed six to

eight weeks post-operatively, with no longer-term testing undertaken. A later cognitive

assessment may be clinically relevant, but cognitive dysfunction in the post-operative period

may predict further decline over the next five years30, 34.

Significant post-operative decline affecting older patients was observed in 6/7 domains

17

assessed, while persistent decline at follow-up was demonstrated in three domains. Of

concern, even younger patients had significant post-operative decline in manual dexterity,

with persistent decline in the use of their dominant hand at follow-up. Silent cerebral

infarction and cognitive decline has implications for patients and should inform the clinical

decision-making and consent process for TEVAR. Cerebral embolization needs to be

evaluated further with differentiation between gaseous and solid emboli, and the efficacy of

pharmacological and interventional strategies to reduce neurological injury in TEVAR and

other endovascular procedures should be determined. Recent randomised control trials such

as CLEAN-TAVI and MISTRAL-C which used the SentinelTM cerebral protection system in

TAVI have shown a significant decrease in number and volume of new silent infarcts, with

improved neurological outcome35,36. Robotic navigation has been shown to reduce contact

with the aortic arch wall during TAVI, and cause significantly less cerebral embolization

compared to manual techniques during TEVAR due to the active manuvrability and stability

of the robotic system37,38. A dynamic bubble trap designed to reduce gaseous emboli has been

shown to reduce TCD HITS, and congnitive function three-months post-operatively was

significantly better compared to controls in patients undergoing coronary artery bypass

grafting39. It has also been demonstrated that thoracic endo-grafts release significant amounts

of air during deployment if flushed according to the instructions for use, and carbon dioxide

flushing prior to standard saline flushing significantly reduces the amount of gas released40.

This may have a potential role to reduce air embolism during TEVAR, particularly in more

proximal landing zones41. Aortic atheroma severity can be addressed with pre-operative statin

usage which leads to plaque stabilization and atheroma regression, and smoking cessation is

beneficial for plaque stabilisation42.

Limitations

18

Due to pragmatic issues with conducting clinical research of this nature it was not feasible for

all patients to undergo all three investigations of TCD, MRI and neurocognitive testing.

Follow-up neurocognitive assessment was only performed at the first OPA to coincide with

routine follow-up and thereby limit the number of clinic visits.

ACKNOWLEDGEMENTS

We would like to thank all members of the vascular surgery and anaesthetic team at St

Mary’s Hospital, in particular Mr Michael Jenkins, Ms Celia Riga, Professor Alun Davies

and Dr Nathalie Courtois for their assistance in conducting this study. Special thanks to

Neuroradiologists Dr Siok Lim for setting up the MRI study and Dr Abhinav Singh for

performing the MRI analysis. Thanks to Agnes Kocsis and Aisling Buckley from the Clinical

Health Psychology team for assistance with neurocognitive testing, and Ms Gagandeep

Grover and Dr Kimberley Kok for helping with recruitment of patients.

19

REFERENCES

1. Patterson B, Holt P, Nienaber C, et al. Aortic pathology determines midterm outcome

after endovascular repair of the thoracic aorta: report from the Medtronic Thoracic

Endovascular Registry (MOTHER) database. Circulation. 2013;127:24-32.

2. Scharrer-Pamler R, Kotsis T, Kapfer X, et al. Complications after endovascular

treatment of thoracic aortic aneurysms. J Endovasc Ther. 2003;10:711-718.

3. Kilic A, Sultan IS, Arnaoutakis GJ, et al. Assessment of Thoracic Endografting

Operative Mortality Risk Score: Development and Validation in 2,000 Patients. Ann Thorac

Surg. 2015;100:860-867.

4. Ullery BW, McGarvey M, Cheung AT, et al. Vascular distribution of stroke and its

relationship to perioperative mortality and neurologic outcome after thoracic endovascular

aortic repair. J Vasc Surg. 2012;56:1510-1517.

5. Gutsche JT, Cheung AT, McGarvey ML, et al. Risk factors for perioperative stroke

after thoracic endovascular aortic repair. Ann Thorac Surg. 2007;84:1195-1200.

6. Patterson BO, Holt PJ, Nienaber C, et al. Management of the left subclavian artery

and neurologic complications after thoracic endovascular aortic repair. J Vasc Surg.

2014;60:1491-1497.

20

7. Feezor RJ, Martin TD, Hess PJ, et al. Risk factors for perioperative stroke during

thoracic endovascular aortic repairs (TEVAR). J Endovasc Ther. 2007;14:568-573.

8. Kotelis D, Bischoff MS, Jobst B, et al. Morphological risk factors of stroke during

thoracic endovascular aortic repair. Langenbecks Arch Surg. 2012;397:1267-1273.

9. Bismuth J, Garami Z, Anaya-Ayala JE, et al. Transcranial Doppler findings during

thoracic endovascular aortic repair. J Vasc Surg. 2011;54:364-369.

10. Kim K, Reynolds T, Donayre C, et al. Predictability of cerebral embolization from

aortic arch manipulations during thoracic endovascular repair. Am Surg. 2011;77:1399-1409.

11. Sacco RL, Kasner SE, Broderick JP, et al. An updated definition of stroke for the 21st

century: a statement for healthcare professionals from the American Heart

Association/American Stroke Association. Stroke. 2013;44:2064-2089.

12. Debette S, Beiser A, DeCarli C, et al. Association of MRI markers of vascular brain

injury with incident stroke, mild cognitive impairment, dementia, and mortality: the

Framingham Offspring Study. Stroke. 2010;41:600-606.

13. Wu RH, Feng C, Xu Y, et al. Late-onset depression in the absence of stroke:

associated with silent brain infarctions, microbleeds and lesion locations. Int J Med Sci.

2014;11:587-592.

14. Bendszus M, Stoll G. Silent cerebral ischaemia: hidden fingerprints of invasive

medical procedures. Lancet Neurol. 2006;5:364-372.

15. Seo DH, Yoon SM, Park HR, et al. Thromboembolic event detected by diffusion

weighted magnetic resonance imaging after coil embolization of cerebral aneurysms. J

Cerebrovasc Endovasc Neurosurg. 2014;16:175-183.

16. Omran H, Schmidt H, Hackenbroch M, et al. Silent and apparent cerebral embolism

after retrograde catheterisation of the aortic valve in valvular stenosis: a prospective,

randomised study. Lancet. 2003;361:1241-1246.

21

17. Bonati LH, Jongen LM, Haller S, et al. New ischaemic brain lesions on MRI after

stenting or endarterectomy for symptomatic carotid stenosis: a substudy of the International

Carotid Stenting Study (ICSS). Lancet Neurol. 2010;9:353-362.

18. Astarci P, Glineur D, Kefer J, et al. Magnetic resonance imaging evaluation of

cerebral embolization during percutaneous aortic valve implantation: comparison of

transfemoral and trans-apical approaches using Edwards Sapiens valve. Eur J Cardiothorac

Surg. 2011;40:475-479.

19. Kahlert P, Eggebrecht H, Janosi RA, et al. Silent cerebral ischemia after thoracic

endovascular aortic repair: a neuroimaging study. Ann Thorac Surg. 2014;98:53-58.

20. Katz ES, Tunick PA, Rusinek H, et al. Protruding aortic atheromas predict stroke in

elderly patients undergoing cardiopulmonary bypass: experience with intraoperative

transesophageal echocardiography. J Am Coll Cardiol. 1992;20:70-77.

21. Mitchell RS, Ishimaru S, Ehrlich MP, et al. First International Summit on Thoracic

Aortic Endografting: roundtable on thoracic aortic dissection as an indication for

endografting. J Endovasc Ther. 2002;9 Suppl 2:Ii98-105.

22. Brazzelli M, Sandercock PA, Chappell FM, et al. Magnetic resonance imaging versus

computed tomography for detection of acute vascular lesions in patients presenting with

stroke symptoms. Cochrane Database Syst Rev. 2009:CD007424.

23. Murkin JM, Newman SP, Stump DA, et al. Statement of consensus on assessment of

neurobehavioral outcomes after cardiac surgery. Ann Thorac Surg. 1995;59:1289-1295.

24. Ringelstein EB, Droste DW, Babikian VL, et al. Consensus on microembolus

detection by TCD. International Consensus Group on Microembolus Detection. Stroke.

1998;29:725-729.

25. Basic identification criteria of Doppler microembolic signals. Consensus Committee

of the Ninth International Cerebral Hemodynamic Symposium. Stroke. 1995;26:1123.

22

26. Bendszus M, Koltzenburg M, Bartsch AJ, et al. Heparin and air filters reduce embolic

events caused by intra-arterial cerebral angiography: a prospective, randomized trial.

Circulation. 2004;110:2210-2215.

27. Perera AH, Gibbs, RGJ. The frequency of cerebral embolisation after aortic arch

catheterisation. In: Greenhalgh RM, ed. Vascular and Endovascular Controversies Update, 37

ed. BIBA;2015:53-59.

28. Vermeer SE, Prins ND, den Heijer T, et al. Silent brain infarcts and the risk of

dementia and cognitive decline. N Engl J Med. 2003;348:1215-1222.

29. Monk TG, Weldon BC, Garvan CW, et al. Predictors of cognitive dysfunction after

major noncardiac surgery. Anesthesiology. 2008;108:18-30.

30. Barber PA, Hach S, Tippett LJ, et al. Cerebral ischemic lesions on diffusion-weighted

imaging are associated with neurocognitive decline after cardiac surgery. Stroke.

2008;39:1427-1433.

31. Ghanem A, Kocurek J, Sinning JM, et al. Cognitive trajectory after transcatheter

aortic valve implantation. Circ Cardiovasc Interv. 2013;6:615-24.

32. Lund C, Nes RB, Ugelstad TP, et al. Cerebral emboli during left heart catheterization

may cause acute brain injury. Eur Heart J. 2005;26:1269-1275.

33. Schwarz N, Schoenburg M, Mollmann H, et al. Cognitive decline and ischemic

microlesions after coronary catheterization. A comparison to coronary artery bypass grafting.

Am Heart J. 2011;162:756-763.

34. Newman MF, Kirchner JL, Phillips-Bute B, et al. Longitudinal assessment of

neurocognitive function after coronary-artery bypass surgery. N Engl J Med. 2001;344:395-

402.

35. Linke A, Haussig S, Dwyer M, et al. CLEAN-TAVI: A prospective, randomized trial

of cerebral embolic protection in high-risk patients with aortic stenosis undergoing

23

transcatheter aortic valve replacement. Transcatheter Cardiovascular Therapeutics Annual

Meeting 2014.

36. Van Mieghem NM, van Gils L, Ahmad H, et al. Filter-based cerebral embolic

protection with transcatheter aortic valve implantation: the randomised MISTRAL-C trial.

EuroIntervention. 2016;12:499-507.

37. Rippel RA, Rolls AE, Riga CV, et al. The use of robotic endovascular catheters in the

facilitation of transcatheter aortic valve implantation. Eur J Cardiothorac Surg. 2014;45:836-

41.

38. Perera AH, Riga CV, Monzon L, et al. Robotic Arch Catheter Placement Reduces

Cerebral Embolization During Thoracic Endovascular Aortic Repair (TEVAR). Eur J Vasc

Endovasc Surg. 2017;53:362-9.

39. Gerriets T, Schwarz N, Sammer G, et al. Protecting the brain from gaseous and solid

micro-emboli during coronary artery bypass grafting: A randomized controlled trial. Eur

Heart J. 2010;31:360-8.

40. Rohlffs F, Tsilimparis N, Saleptsis V, et al. Air Embolism During TEVAR: Carbon

Dioxide Flushing Decreases the Amount of Gas Released from Thoracic Stent-Grafts During

Deployment. J Endovasc Ther. 2017;24:84-8.

41. Kolbel T, Rohlffs F, Wipper S, et al. Carbon Dioxide Flushing Technique to Prevent

Cerebral Arterial Air Embolism and Stroke During TEVAR. J Endovasc Ther. 2016;23:393-

5.

42. Yla-Herttuala S, Bentzon JF, Daemen M, et al. Stabilization of atherosclerotic

plaques: an update. Eur Heart J. 2013;34:3251-8.

24

LEGENDS FOR ILLUSTRATIONS

Figure 1. Intra-operative transcranial Doppler at the left and right middle cerebral artery

(MCA) in a 74-year-old male patient undergoing TEVAR and left carotid-subclavian bypass

for repair of thoracic aortic aneurysm. Proximal landing zone at zone 1 using custom-made

stent-graft with proximal scallop for the left common carotid origin.

a. No spontaneous embolization detected during pre-operative monitoring

b. 3 HITS at left MCA and 8 HITS at right MCA (red arrows) during wire/catheter

exchange

25

c. Multiple HITS bilaterally during stent-graft deployment

26

Figure 2. Rate of embolization with each procedural phase of TEVAR

27

n=41 (1 patient excluded from graph as outlier for scaling purposes but not excluded from

statistical analysis)

Wire and catheter manipulation: median 20 (IQR 10-41)

Stent related (manipulation and deployment): median 62 (IQR 22-163)

Contrast injection total: median 61 (IQR 18-117)

p=0.001

Figure 3. Embolization rates with patient and procedural factors

a. Atheroma grade

28

n=41(1 patient excluded from graph as outlier for scaling purposes but not excluded

from statistical analysis)

Grade 1-3: median 100 (IQR 52-211)

Grade 4 &5: median 207 (124-450)

p=0.042

b. MRI outcome

29

n=42

NAD- no lesions detected on MRI: median 97 (IQR 29-236)

SCI- silent cerebral infarction on MRI: median 195 (IQR 138-280)

Stroke- Stoke clinically, confirmed on MRI: median 299 (IQR 146-908)

p=0.498

Figure 4. Pre- and post-operative cerebral DW-MRI of 68-year old male patient undergoing

arch hybrid for repair of thoracic aortic aneurysm. Multiple new silent cerebral infarcts are

seen post-operatively (red arrows).

30

a. Pre-operative DW-MRI

b. Post-operative DW-MRI

Figure 5. Neurocognitive outcomes

a. Results of controlled oral word association test; executive cognitive dysfunction

31

SCI age </=69 years group: p=0.779

SCI age >70 years group: p=0.009

Pre-operative to post-operative p=0.027

Pre-operative to OPA p=0.050

b. Results of trails A; visual search and motor skills

32

The higher the score, the longer the time taken to complete task

SCI age </=69 years group: p=0.050

SCI age >70 years group: p=0.115

Pre-operative to post-operative p=0.116

Pre-operative to OPA p=0.028

c. Results of pegboard test dominant (right) hand

33

The higher the score, the longer the time taken to complete task

SCI age </=69 years group: p=0.018

Pre-operative to post-operative p=0.046

Pre-operative op to OPA p=0.028

SCI age >70 years group: p=0.011

Pre-operative op to post-operative p=0.028

Pre-operative to OPA p=0.310

Table 1. Patient demographics and procedural characteristics

34

Variables TEVAR group

(n=52)

Control group

(n=24)

Age, years

Male sex

Aetiology

Atherosclerotic aneurysm

Dissection

(Connective tissue disease)

Anastomotic pseudoaneurysm following open aortic

coarctation repair

Penetrating aortic ulcer

Mycotic aneurysm

Pseudoaneruysm following aortic transection

Pathology

Aneurysm

Crawford type I

Crawford type II

Crawford type III

Aortic arch aneurysm

Descending thoracic aortic aneurysm

Saccular aneurysm

66 (53.5-77)

32 (62%)

22

11

2

8

7

2

2

46 (88%)

1

7

3

11

11

7

79 (73-81)

22 (92%)

24

0

0

0

0

0

0

24 (100%)

0

0

0

0

0

0

35

Abdominal aortic aneurysm

Endoleak following previous TEVAR

Chronic type B dissection + aneurysmal dilatation

Penetrating aortic ulcer + aneurysm

Acute aortic syndrome (type B dissection/IMH)

Aneurysm size, cm

Atherosclerotic

Saccular

Post-dissection

Patient co-morbidities

Hypertension

Hypercholesterolemia

Diabetes mellitus

Smoking history

IHD

Respiratory disease

Cancer

Previous cardiovascular surgery

Chronic kidney disease

Medication

0

6

4

6

6 (12%)

6.7 (6.0-8.1)

4.8 (3.5-5.4)

7.3 (6.5-10.3)

34 (65%)

27 (52%)

4 (8%)

34 (65%)

7 (14%)

15 (29%)

8 (15%)

28 (54%)

7 (13%)

24

0

0

0

0

5.9 (5.7-6.1)

N/A

N/A

17 (71%)

13 (54%)

2 (8%)

15 (63%)

10 (42%)

3 (13%)

0

5 (21%)

3 (13%)

36

Single antiplatelet

Warfarin

Statin

Significant (>50%) carotid stenosis on duplex

Procedures

TEVAR with no adjunct procedures

TEVAR with L carotid-subclavian bypass

TEVAR with LSCA coverage

Arch Hybrid

Visceral Hybrid

TEVAR with visceral vessel fenestrations

Arch branched graft

Infra-renal EVAR

FEVAR with visceral vessel fenestrations

Urgency

Elective

Stent-grafts

Gore cTag

Valiant Medtronic

Custom-made Bolton Relay proximal scallop

Custom-made Bolton Relay arch branched graft

15 (29%)

3 (6%)

27 (52%)

0

20 (38%)

14 (27%)

4 (8%)

5 (10%)

5 (10%)

3 (6%)

1 (2%)

0

0

34 (65%)

23 (44%)

11 (21%)

11 (21%)

1 (2%)

10 (42%)

2 (8%)

13 (54%)

N/A

0

0

0

0

0

0

0

19 (79%)

5 (21%)

24 (100%)

0

0

0

0

37

Custom-made Cook with visceral fenestrations

Medtronic Endurant II

Gore Excluder

Proximal landing zone

Zone 0

Zone 1

Zone 2

Zone 3

Zone 4

Abdominal aorta

Length of stay, days

6 (12%)

0

0

3 (6%)

4 (8%)

25 (48%)

14 (27%)

6 (12%)

0

8 (4-14)

5 (21%)

15 (62%)

4 (17%)

0

0

0

0

0

24 (100%)

4 (3-6)

Continuous data are shown as median (interquartile range) and categorical data are shown as

number (percentage)

Table 2. Morbidity and mortality outcomes

38

Re-intervention

Open repair of femoral pseudoaneurysm (following

failure of percutaneous closure device)

Late type Ib endoleak

Type III endoleak

Extension of penetrating aortic ulcer requiring distal stent

extension

Iliac artery rupture

1 (2%)

3 (6%)

2 (4%)

1 (2%)

1 (2%)

Morbidity

Dural leak with spinal drain (resolved spontaneously)

Brachial plexus neuropraxia following carotid-subclavian

bypass (resolved with physiotherapy

Paraplegia following visceral hybrid repair

Clinical stroke

1 (2%)

1 (2%)

2 (4%)

4 (7.7%)

30-day mortality

Myocardial infarction and cardiac arrest following arch

hybrid

sepsis secondary to bowel ischemia and cardiac arrest

following visceral hybrid

1 (2%)

1 (2%)

Table 3. MRI outcomes

Patient MRI Clinical Number Laterality Cerebral

39

number outcome outcome of lesions of lesions circulation

1-6 Nil Nil 0 N/A N/A

7 Stroke MRS 5 Multiple Bilateral Posterior

8 Stroke MRS 2

(Expressive

dysphasia,

resolved

completely

with rehab)

Large

single

lesion

Left Anterior

9 Stroke MRS 5 Multiple Bilateral Anterior and

posterior

10 Stroke MRS 2

(Right hand

weakness.

resolved

completely

with rehab)

Multiple Left Anterior and

border-zone

11DG SCI Nil 1 Left Anterior

12DM SCI Nil 2 Bilateral Anterior and

border-zone

13AC SCI Nil 2 Bilateral Posterior and

border-zone

14OM SCI Nil 1 Left Posterior

15SB SCI Nil 2 Bilateral Border-zone

16QA SCI Nil 4 Bilateral Anterior and

40

posterior

17JP SCI Nil 1 Left Posterior

18AH SCI Nil 1 Left Posterior

19RL SCI Nil 20 Bilateral Anterior,

posterior and

border-zone

20DL SCI Nil 14 Bilateral Posterior and

border-zone

21MB SCI Nil 15 Bilateral Posterior and

border-zone

22RR SCI Nil 25 Bilateral Anterior,

posterior and

border-zone

23RB SCI Nil 1 Left Border-zone

24SP SCI Nil 3 Bilateral Anterior and

posterior

25BL SCI Nil 1 Left Border-zone

26GD SCI Nil 13 Bilateral Posterior

27WE SCI Nil 5 Left Anterior

28MC SCI Nil 1 Left Anterior

29ME SCI Nil 3 Bilateral Posterior

30 SCI Nil 2 Left Anterior and

border-zone

31 SCI Nil 1 Left Border-zone

41

N/A Non-applicable

MRS Modified Rankin Scale 0-6

1 No significant disability (able to carry out all usual activities)

2 Slight disability (unable to carry out all previous activity but able to look after own affairs

without assistance)

3 Moderate disability (requiring some help but able to walk without assistance)

4 Moderately severe disability (unable to walk or attend to bodily needs without assistance)

5 Severe disability (bedridden, incontinent, requiring constant care and attention)

6 Dead

Table 4. Predictors of TCD-detected cerebral emboli

Risk factor ß Coefficient Standard error p value

LSCA

revascularisation

0.336 130.28 0.030

42

(bypass and scallop

graft)

LSCA bypass 0.423 132.62 0.005

Proximal landing

zone

0.451 89.10 0.003

PLZ 0 and 1 0.504 170.57 0.001

Procedure 0.356 61.38 0.021

Arch hybrid 0.514 182.96 0.000

43