Using machine learning to predict carotid artery symptoms from CT angiography: A radiomics and deep learning approach.

Purpose: To assess radiomics and deep learning (DL) methods in identifying symptomatic Carotid Artery Disease (CAD) from carotid CT angiography (CTA) images. We further compare the performance of these novel methods to the conventional calcium score. Methods: Carotid CT angiography (CTA) images from symptomatic patients (ischaemic stroke/transient ischaemic attack within the last 3 months) and asymptomatic patients were analysed. Carotid arteries were classified into culprit, non-culprit and asymptomatic. The calcium score was assessed using the Agatston method. 93 radiomic features were extracted from regions-of-interest drawn on 14 consecutive CTA slices. For DL, convolutional neural networks (CNNs) with and without transfer learning were trained directly on CTA slices. Predictive performance was assessed over 5-fold cross validated AUC scores. SHAP and GRAD-CAM algorithms were used for explainability. Results: 132 carotid arteries were analysed (41 culprit, 41 non-culprit, and 50 asymptomatic). For asymptomatic vs symptomatic arteries, radiomics attained a mean AUC of 0.96(± 0.02), followed by DL 0.86(± 0.06) and then calcium 0.79(± 0.08). For culprit vs non-culprit arteries, radiomics achieved a mean AUC of 0.75(± 0.09), followed by DL 0.67(± 0.10) and then calcium 0.60(± 0.02). For multi-class classification, the mean AUCs were 0.95(± 0.07), 0.79(± 0.05), and 0.71(± 0.07) for radiomics, DL and calcium, respectively. Explainability revealed consistent patterns in the most important radiomic features. Conclusions: Our study highlights the potential of novel image analysis techniques in extracting quantitative information beyond calcification in the identification of CAD. Though further work is required, the transition of these novel techniques into clinical practice may eventually facilitate better stroke risk stratification.

End of Life Care for Unplanned Vascular Admissions.

BACKGROUND: Unplanned vascular admissions have a high mortality. Previous studies have indicated that end of life care (EoLC) among this group of patients is low but there exist limited data on EoLC in the United Kingdom. The aim of this study was to evaluate the quality and predictors of EoLC for unplanned vascular admissions to a tertiary center in the United Kingdom. METHODS: This was a retrospective single-center cohort study of unplanned vascular surgery admissions from August 1, 2019 to January 22, 2020. Data on patient demographics, markers of quality of palliative care, mortality, and cause of death of unplanned admission to the vascular surgery department were collected from hospital and general practitioner records and evaluated against EoLC to evaluate predictors and efficacy of EoLC. Quality of palliative care markers included documentation of preferred place of death and care priorities, time spent in hospital and the intensive care unit toward the end of life, and realization of documented care objectives. EoLC input was defined as a dedicated palliative care consultation (PCC) by a palliative care professional, medical doctor, surgeon, or advanced care practitioner. We also conducted a subgroup analysis of patients within this group with chronic limb-threatening ischemia (CLTI), diabetic foot, and ruptured aortic aneurysms, as all patients in this group should be offered EoLC according to international guidelines. RESULTS: One-hundred and fifty patients were included. Median age at presentation was 70.5 years, and the cohort consisted of mostly men (72%). CLTI (31%) was the most common reason for admission. Surgical intervention was carried out in 60% of patients. Two-year mortality was 36%, and pneumonia (22%) was the most common cause of death. Seven percent of patients received PCC, which occurred a median of 10 days before death. Only a minority of patients had preferred place of care/death (14%), care priorities (37%), and family involvement during advance care planning (17%) documented in their notes; 29% of patients had Recommended Summary Plan for Emergency Care and Treatment forms in place. A diagnosis of left ventricular systolic dysfunction, chronic kidney disease, and increasing age predicted Recommended Summary Plan for Emergency Care and Treatment form completion. Patients with PCC were more likely to have advance care planning, but this did not translate into improvements in the other markers of quality of palliative and, consequently, EoLC. CONCLUSIONS: EoLC was insufficient and of low quality despite a high mortality in this group. Clinical guidelines and pathways are needed to ensure these patients are considered for EoLC and those with CLTI, diabetic foot sepsis or ruptured abdominal aortic aneurysms are offered it by default. Further research is needed to help identify vascular patients who would benefit from EoLC earlier to improve quality at end of life.

Assessment of internet-based information on statin therapy.

AIMS: The use of statin therapy is deemed to be controversial by mainstream media. Patients increasingly source medical information from the internet, and the use of statins is no exception. This study aims to determine the quality and educational content of statin-focused information on the internet and YouTube. METHODS AND RESULTS: 'Statin' was searched on Google, Yahoo!, Bing, and YouTube. The first 50 results obtained from each search engine and the first 20 YouTube videos were screened by two assessors. Websites were assessed using the Flesch Reading Ease (FRE) score, University of Michigan Consumer Health Website Evaluation Checklist, and a customized scoring system evaluating statin-focused content for quality. Videos were scored using the Journal of the American Medical Association (JAMA) benchmark criteria, Global Quality Score (GQS), and the customized scoring system. Websites scored a median FRE score of 57.5 [interquartile range (IQR) 52.1-62.3], median Michigan score of 36 (IQR 32-41.5), and median content score of 5 (IQR 3.75-7). Good interobserver agreement was demonstrated [Michigan score interobserver coefficient correlation (ICC) = 0.968; content score ICC = 0.944]. Videos scored a median JAMA score of 2, median GQS score of 2.5, and median content score of 2.5. Good interobserver agreement was demonstrated (JAMA ICC = 0.746; GQS ICC = 0.874; content score ICC = 0.946). CONCLUSION: Quality and readability of statin-focused online information are poor. Healthcare professionals should be aware of the limitations of the current available sources and design online resources that are accurate and patient-friendly.

Assessment of Available Online Website and YouTube Resources for Patients with Abdominal Aortic Aneurysms.

BACKGROUND: Over the last decade, patients have displayed a greater tendency to search for online information related to their health before seeking advice from a clinician. This study aims to determine the current quality and educational content of online patient information for abdominal aortic aneurysms (AAAs). METHODS: In March 2022, the 3 most popular search engines by market shares (Google, Yahoo!, and Bing) and the video platform YouTube were interrogated for the term "abdominal aortic aneurysm". Validated scoring tools were used to assess quality and readability of the top 50 results for each search engine and to evaluate reliability and educational quality of the first 20 YouTube videos returned by the search. A custom-made scoring system was used to assess content. RESULTS: Forty-five unique websites were analysed, 29% of which held Health on the Net certification. Median Flesch-Kincaid Reading Ease (interquartile range [IQR]) was 56.4 (50.4-62.75), with the average website falling under the "difficult to read" category. Median Michigan score (IQR) was 38.5 (32-43.5), reflecting "weak" quality. Websites with a higher content-specific score had a significantly higher median Michigan score. Sixty percent of websites discussed benefits and risks related to AAA treatment, and only 31% discussed advantages and disadvantages of open versus endovascular treatment. No websites mentioned the volume-outcome relationship in aneurysm surgery. Eight unique YouTube videos were assessed. Median Journal of the American Medical Association score (IQR) was 2 (2-2.25). Median Global Quality Score score (IQR) was 3 (2-4). Median content score was 1 (0-2). CONCLUSIONS: The current average online information on AAA is of 'weak' quality and 'difficult' (i.e., above the standard reading ability of a 13- to 15-year-old) readability. Healthcare providers should focus on the provision of better AAA-focused patient information (e.g., appropriately referenced, regularly reviewed, and limiting advertisements where possible). The involvement of patient advisory groups during resource development is highly recommended.

Analyzing Sex Differences in Intensity of Cardiovascular Disease Prevention Medications in Patients With Abdominal Aortic Aneurysms-A Single-Center Cross-Sectional Study.

BACKGROUND: Patients with abdominal aortic aneurysm (AAA) are at a significant risk of cardiovascular events, similar to that of patients who have already experienced a major cardiac event. The European Society for Vascular Society AAA guidelines suggest that antiplatelet therapy and lipid-lowering therapy (LLT) should be considered in all patients with AAA. This study explores the overall prevalence and intensity of antithrombotic therapy and LLT, and lipid profile monitoring in a single center AAA surveillance cohort alongside any sex differences. METHODS: This was a retrospective, single center, cross-sectional study of 614 patients enrolled in the AAA surveillance program of a tertiary vascular surgery unit. All patients undergoing at least 1 surveillance scan from January 1, 2018, to December 31, 2020, were assessed. Electronic hospital records linked to real-time primary care records were interrogated for data on demographics, comorbidities, antiplatelet and LLT prescriptions, and serum cholesterol laboratory results. An analysis of covariance test was used to account for the effects of confounding comorbidities. RESULTS: Twenty-one percent of patients were not on antithrombotic therapy, and 20% of patients were not on LLT which reflects a group of patients receiving sub-optimal clinical care. In total, 47% of the cohort were on low/moderate intensity statin therapy which reflects a group of patients where care can be improved upon. Female sex was independently associated with a reduced likelihood of being prescribed LLT (P = 0.008, eta squared (ηp2) = 0.012, small effect size) but not antithrombotic therapy (P = 0.202). Fewer women underwent low-density lipoprotein cholesterol (LDL-C) monitoring (mean difference 9%, P = 0.040) and achieved the European Society of Cardiology-European Atherosclerosis Society- LDL-C target of <1.4 mmol/L (mean difference 9%, P = 0.040). CONCLUSIONS: Overall, there is room for improvement in these aspects of cardiovascular risk prevention for both sexes. Sex differences in the prescription of LLT, the prevalence of lipid profile monitoring, and likelihood of achieving LDL-C targets exist among patients with AAA, with a lower prevalence in women.

Comparing the Benefit of Duplex Ultrasound Surveillance Following Both Infrainguinal Bypass Surgery and Stenting for Femoro-Popliteal Disease.

OBJECTIVE: Duplex ultrasound surveillance (DUS) is commonly used following infrainguinal vein bypass. The role of DUS following endovascular revascularisation is as yet unclear. This study focuses on the role of DUS in a contemporary group of patients undergoing infrainguinal bypass or stent insertion. METHODS: All patients undergoing either an infrainguinal vein graft bypass or stent insertion into the femoro-popliteal segment (November 2014 - January 2017) were identified. Patients were followed up for 2 years. Data on entry into DUS, pre-operative characteristics, adjunctive pharmacotherapy and reintervention were collated. The primary outcomes were major lower limb amputation and mortality at 2 years post revascularisation. RESULTS: One hundred and thirty-five patients underwent infrainguinal vein bypass and 100 patients underwent stent insertion. 107 patients in the bypass cohort and 58 patients in the stent cohort entered DUS. For the bypass cohort, entering DUS was associated with a lower mortality rate (P = .003) but was not associated with an improvement in amputation rates. The odds ratio of major amputation or mortality was greater in the no surveillance group (4.58, 95% CI: 1.855 - 11.364). In the stent cohort, DUS was not associated with a significant improvement in either major amputation or death (odds ratio 2.13 (95% CI 0.903 - 5.051; P = .081). CONCLUSION: DUS was associated with improved survival rates in patients undergoing lower limb bypass but had no benefit in those patients undergoing stent insertion. The role of DUS following stent insertion in the femoropopliteal segment needs to be better defined.

Frailty predicts poor longer-term outcomes in patients following lower limb open surgical revascularization.

BACKGROUND: Frailty in vascular surgery patients is increasingly recognized as a marker of poor outcome. This provides particular challenges for patients with lower limb peripheral arterial disease who require surgical revascularization. This study aimed to assess the impact of frailty on short- and long-term outcome in this specific patient group using a specialty specific frailty score. METHODS: Patients undergoing open surgical revascularization for chronic limb ischemia (January 2015-December 2016) were assessed. Demographics, mode of admission, diagnosis, and site of surgery were recorded alongside a variety of frailty-specific characteristics. We calculated the previously validated Addenbrookes Vascular Frailty Score (AVFS) and Long AVFS (LAVFS). Primary outcome was 3-year mortality. RESULTS: Two hundred and sixty-one patients (75% men, median age 69 years) were studied. The median length of stay was 6 days with a 3-year mortality of 23%. The predictive power of vascular frailty scores showed that for 3-year mortality, area under the receiver operator curve values (AUROC) were specific for both the AVFS score (AUROC: 0.724, 95% CI: 0.654-0.794) and LAVFS Score (AUROC: 0.741, 95%CI: 0.670-0.813). Furthermore, the cumulative AVFS and LAVFS scores both predicted mortality over the follow-up period (P=0.0001) with increased mortality among patients with higher scores. CONCLUSIONS: Incremental worsening of frailty, determined using a specialty specific frailty score, predicts mortality risk in patients undergoing lower limb surgical revascularization.

Intravascular Fluorescence Molecular Imaging of Atherosclerosis.

Optical molecular imaging using near-infrared fluorescence (NIRF) light is an emerging high-resolution imaging approach to image a wide range of molecular and cellular species in vivo. Imaging using NIR wavelengths (650-900 nm) enables deeper photon penetration into tissue and reduced tissue autofluorescence, resulting in higher sensitivity to detect exogenously administered NIR fluorophores (injectable molecular imaging agents). Greater imaging depth of several centimeters is further achievable in the NIR window as blood absorption is as an order of magnitude lower than in the visible range. Furthermore, as optical imaging is routinely performed in the cardiac catheterization laboratory (e.g., optical coherence tomography), intravascular NIRF offers a promising translational approach for clinical coronary and peripheral arterial imaging. To this point, the first human intravascular NIRF imaging study recently demonstrated the ability to detect NIR autofluorescence in patients with coronary atherosclerosis. This study provides a foundation for targeted intravascular NIRF molecular imaging studies in coronary patients. In this chapter, we detail system engineering, imaging agents and translational applications of intravascular NIRF molecular imaging.

Systematic review and narrative synthesis of surveillance practices after endovascular intervention for lower limb peripheral arterial disease.

OBJECTIVE: The optimal timing and modality of surveillance after endovascular intervention for peripheral arterial disease is controversial, and no randomized trial to assess the value of peripheral endovascular intervention has ever been performed. The aim of this systematic review was to examine the practice of surveillance after peripheral endovascular intervention in randomized trials. METHODS: We used the Medline, Embase, Cochrane Library, and WHO trial registry databases in this systematic review of the literature to capture surveillance strategies used in randomized trials comparing endovascular interventions. Surveillance protocols were assessed for completeness, modalities used, duration, and intensity. RESULTS: Ninety-six different surveillance protocols were reported in 103 trials comparing endovascular interventions. Protocol specification was incomplete in 32% of trials. The majority of trials used multiple surveillance modalities (mean of 3.46 modalities), most commonly clinical examination (96%), ankle-brachial index (80%), duplex ultrasound examination (75%), and digital subtraction angiography (51%). Trials involving infrapopliteal lesions used more angiographic surveillance than trials with femoropopliteal lesions (P = .006). The median number of surveillance visits in the first 12 months after intervention was three and the mean surveillance duration was 21 months. Trials treating infrapopliteal vessels had a higher surveillance intensity compared with those treating femoropopliteal lesions in the first 12 months after endovascular intervention (mean 5 vs 3 surveillance visits; P = .017). Trials with drug-eluting devices had longer surveillance duration compared with those without (mean 26 vs 19 months; P = .020). CONCLUSIONS: There is a high level of variation in the modality, duration, and intensity of surveillance protocols used in randomized trials comparing different types of peripheral endovascular arterial intervention. Further research is required to determine the value and impact of postprocedural surveillance on patient outcomes.

Effects of Statin Therapy and Dose on Cardiovascular and Limb Outcomes in Peripheral Arterial Disease: A Systematic Review and Meta-analysis.

OBJECTIVE: Statin therapy is indicated in patients with peripheral arterial disease (PAD). National Institute for Health and Care Excellence guidelines suggest the use of "high intensity" statins, although evidence with PAD specific data are lacking. The effect of statin therapy and dose on outcomes in PAD is investigated. DATA SOURCES: Studies measuring statin use in PAD patients and outcomes were identified based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) criteria. The EMBASE and MEDLINE databases were interrogated from January 1957 until February 2020. Twenty-two observational cohort studies and two randomised control trials were included (n = 268 611). REVIEW METHODS: Pooled estimates of dichotomous outcome data were calculated using the odds/hazard ratios (OR/HR) and 95% confidence interval (CI). Meta-analysis was conducted using the inverse variance or Mantel-Haenszel method. Outcomes included all cause mortality (ACM), cardiovascular mortality (CVM), major adverse cardiac events (MACE), and amputation. Subgroup analysis was performed on studies comparing patients taking high dose vs. combined low and moderate doses of statins. The GRADE criteria assessed the quality of evidence for outcomes. RESULTS: Statin therapy (vs. no statins) was significantly protective for ACM: OR 0.68 (95% CI 0.60 - 0.76) (number needed to treat [NNT] = 48), HR 0.74 (95% CI 0.70 - 0.78) (NNT = 10 - 91); MACE: OR 0.84 (95% CI 0.78 - 0.92) (NNT = 53), HR 0.78 (95% CI 0.65 - 0.93) (NNT = 167); and amputations: OR 0.59 (95% CI 0.33 - 1.07) (NNT = 333), HR 0.74 (95% CI 0.62 - 0.89) (NNT = 50). High doses of statins (vs. combined low and moderate doses) were significantly better protective against ACM OR 0.69 (95% CI 0.43 - 1.09) (NNT = 17), HR 0.74 (95% CI 0.62 - 0.89) (NNT = 16 - 200) but work less significantly for MACE OR 0.77 (95% CI 0.49 - 1.21) (NNT = 25). Amputations were less frequent in patients on high doses HR 0.78 (95% CI 0.69 - 0.90) (NNT = 53 - 1 000). CONCLUSION: Higher dosing of statins confers a significant improvement in patient outcomes, especially ACM and amputations, although the quality of the evidence was variable. Such findings require confirmation in larger, PAD specific trials.

An Assessment of Available Information on the World Wide Web for Patients with Lower Limb Arterial Disease.

OBJECTIVE: The quality of patient information relating to intermittent claudication (IC) and peripheral arterial disease (PAD) on the World Wide Web was assessed. METHODS: The quality of websites and YouTube videos was assessed using the search terms "intermittent claudication" and "peripheral arterial disease". The first 50 hits screened for each search term from the three largest search engines by market share, and the first 20 videos from YouTube were screened. Website quality was scored using the University of Michigan Consumer Health Website tool (maximum score 80). Readability was calculated using the Flesch Reading Ease (FRE) score (maximum score 100). Videos were classified by content and upload source. Video reliability was assessed using the JAMA benchmark criteria. Video educational content was assessed using the Global Quality Score (GQS). Subjective content assessment was undertaken. RESULTS: Seventy-six websites were analysed. The majority of websites for both IC (51.7%) and PAD (72.3%) were rated as weak. The median Michigan score for IC (27; interquartile range [IQR] 15, 32.5) was lower that the score for PAD (31; IQR 25.5, 38.8; p = .030). The majority of websites for both IC (69%) and PAD (68.1%) were rated as requiring an above average reading level. The overall median FRE score was 55.9 (IQR 46.6, 60.6) for IC and 55.3 (IQR 44.6, 59.3) for PAD. Twenty-two videos were analysed. Of the 14 videos evaluated as part of the PAD search, the median JAMA score was 2 (2 - 3), the median GQS score was 3 (3 - 3) and the evaluation of content score was 8.5 (7.25 - 11.5). The equivalent scores for the IC search were 2 (2 - 2), 3 (3 - 4), and 5.5 (5 - 8). CONCLUSION: The educational quality and reliability of information both in written and video form on the internet is low. Attention needs to focus on improving the quality of all forms of information delivery to allow proper advocacy for patients.

Assessing robustness of carotid artery CT angiography radiomics in the identification of culprit lesions in cerebrovascular events.

Radiomics, quantitative feature extraction from radiological images, can improve disease diagnosis and prognostication. However, radiomic features are susceptible to image acquisition and segmentation variability. Ideally, only features robust to these variations would be incorporated into predictive models, for good generalisability. We extracted 93 radiomic features from carotid artery computed tomography angiograms of 41 patients with cerebrovascular events. We tested feature robustness to region-of-interest perturbations, image pre-processing settings and quantisation methods using both single- and multi-slice approaches. We assessed the ability of the most robust features to identify culprit and non-culprit arteries using several machine learning algorithms and report the average area under the curve (AUC) from five-fold cross validation. Multi-slice features were superior to single for producing robust radiomic features (67 vs. 61). The optimal image quantisation method used bin widths of 25 or 30. Incorporating our top 10 non-redundant robust radiomics features into ElasticNet achieved an AUC of 0.73 and accuracy of 69% (compared to carotid calcification alone [AUC: 0.44, accuracy: 46%]). Our results provide key information for introducing carotid CT radiomics into clinical practice. If validated prospectively, our robust carotid radiomic set could improve stroke prediction and target therapies to those at highest risk.

Editor's Choice - Systematic Review and Meta-Analysis of Wound Adjuncts for the Prevention of Groin Wound Surgical Site Infection in Arterial Surgery.

OBJECTIVE: Groin incision surgical site infections (SSIs) following arterial surgery are common and are a source of considerable morbidity. This review evaluates interventions and adjuncts delivered immediately before, during, or after skin closure, to prevent SSIs in patients undergoing arterial interventions involving a groin incision. DATA SOURCES: MEDLINE, EMBASE, and CENTRAL databases were searched. REVIEW METHODS: This review was undertaken according to established international reporting guidelines and was registered prospectively with the International prospective register of systematic reviews (CRD42020185170). The MEDLINE, EMBASE, and CENTRAL databases were searched using pre-defined search terms without date restriction. Randomised controlled trials (RCTs) and observational studies recruiting patients with non-infected groin incisions for arterial exposure were included; SSI rates and other outcomes were captured. Interventions reported in two or more studies were subjected to meta-analysis. RESULTS: The search identified 1 532 articles. Seventeen RCTs and seven observational studies, reporting on 3 747 patients undergoing 4 130 groin incisions were included. A total of seven interventions and nine outcomes were reported upon. Prophylactic closed incision negative pressure wound therapy (ciNPWT) reduced groin SSIs compared with standard dressings (odds ratio [OR] 0.34, 95% CI 0.23 - 0.51; p < .001, GRADE strength of evidence: moderate). Local antibiotics did not reduce groin SSIs (OR 0.60 95% CI 0.30 - 1.21 p = .15, GRADE strength: low). Subcuticular sutures (vs. transdermal sutures or clips) reduced groin SSI rates (OR 0.33, 95% CI 0.17 - 0.65, p = .001, GRADE strength: low). Wound drains, platelet rich plasma, fibrin glue, and silver alginate dressings did not show any significant effect on SSI rates. CONCLUSION: There is evidence that ciNPWT and subcuticular sutures reduce groin SSI in patients undergoing arterial vascular interventions involving a groin incision. Local antibiotics did not reduce groin wound SSI, although the strength of this evidence is lower. No other interventions demonstrated a significant effect.

Patient and Institutional Costs of Failure of Angioplasty of the Superficial Femoral Artery.

BACKGROUND: Debate surrounds the optimal management of superficial femoral artery (SFA) disease. Randomized trial data rarely reflect real world findings, specifically the consequences to the patient of angioplasty failure. We observed the effect of a failed SFA angioplasty on the need for repeated clinic visits, hospital readmissions, imaging requirements, and reinterventions. METHODS: We reviewed a consecutive series of 148 patients (94 men, median age 72 years) undergoing solely SFA angioplasty over a 2-year period. Patient preangioplasty demographics and 2-year post-PTA follow-up data were collated, including hospital attendances (inpatient/outpatient), further imaging (including radiation exposure) and revascularization attempts. We defined "failed angioplasty" as presence of clinical symptoms with radiological evidence of significant restenosis after an initial successful primary SFA angioplasty. RESULTS: Fifty-four patients represented with a failed angioplasty (median time of 4 months after index PTA). In this group, failure of index angioplasty resulted in a further 185 restenosis-related clinic visits and a total of 537 bed days of inpatient stay. This group underwent a further 149 imaging events and required a further 34 endovascular revascularization procedures and 12 infrainguinal bypass procedures. These interventions and investigations corresponded to overall effective radiation dose across all patients of 190.69 mSv. Of the cohort of 99 patients who did not have a "failed angioplasty," they required 100 clinic visits, 21 further scans (total radiation dose 6.42 mSv), and 36 bed days of inpatient admission. CONCLUSIONS: Failed angioplasty results in significant additional consequences for patients and health-care systems. Further work should focus on refining decision-making, providing the right procedure to the right patient at the right time.

Progressive Device Failure at Long Term Follow Up of the Nellix EndoVascular Aneurysm Sealing (EVAS) System.

OBJECTIVE: High rates of midterm failure of the Nellix EndoVascular Aneurysm Sealing (EVAS) System resulted in device withdrawal from the UK market. The study aim was to report long term Nellix EVAS outcomes and management of a failing device. METHODS: A retrospective review of EVAS procedures at a tertiary unit was performed. Device failure was defined as a triad of stent migration, stent separation, and secondary sac expansion, or any intervention for type 1 endoleak, device rupture, or explant. RESULTS: 161 (male n = 140, female n = 21) patients with a median follow up of 6.0 (IQR 5.0-6.6) years were included. Freedom from all cause mortality estimate at six years was 41.5%. There were 70 (43.5%) device failures with a freedom from device failure estimate at six years of 32.3%. Failure was the result of sac expansion (n = 41), caudal stent migration (n = 36), stent separation (n = 26), and secondary AAA rupture (n = 15). A substantial number of type 1 endoleaks was present (1a n = 33, 1b n = 11), but the type 2 endoleak rate was low at 3.7%. Some 36 (22.4%) patients required re-intervention. Twenty-one patients underwent explant with no 30 day deaths. Six patients underwent Nellix-in-Nellix application (NINA) with one early death from bowel ischaemia and one patient who died later from non-aneurysm related cause. Two NINA patients have ongoing sac expansion and two have had thrombosis of a Nellix limb or visceral stent. Proximal embolisation was only successful in one of six cases. CONCLUSION: The long term failure rate of Nellix EVAS is high. All patients with a device must be informed and be enrolled in enhanced surveillance. EVAS explant is an acceptable technique with favourable outcomes. Management by open explant, if the patient is fit, should be considered early and offered to those with device failure.

The Effect of Frailty on Long Term Outcomes in Vascular Surgical Patients.

OBJECTIVE: Frailty is a multidimensional vulnerability due to age associated decline. The impact of frailty on long term outcomes was assessed in a cohort of vascular surgical patients. METHODS: Patients aged over 65 years with a length of stay greater than two days admitted to a tertiary vascular unit over a single calendar year were included. Demographics, mode of admission, and diagnosis were recorded alongside a variety of frailty specific characteristics. Using the previously developed Addenbrookes Vascular Frailty Score (AVFS - 6 point score: anaemia on admission, lack of independent mobility, polypharmacy, Waterlow score > 13, depression, and emergency admission) the effect of frailty on five year mortality and re-admission rates was assessed using multivariable regression techniques. The AVFS was further refined to assess longer term outcomes. RESULTS: In total, 410 patients (median age 77 years) were included and followed up until death or five years from the index admission. One hundred and thirty-four were treated for aortic aneurysm, 75 and 96 for acute and chronic limb ischaemia respectively, 52 for carotid disease, and 53 for other pathologies. The in hospital mortality rate was 3.6%. The one, three, and five year survival rates were 83%, 70% and 59%; and the one, three, and five year re-admission free survival rates were 47%, 29%, and 22% respectively. Independent predictors of five year mortality were age, lack of independent mobility, high Charlson score, polypharmacy, evidence of malnutrition, and emergency admission (p < .010 for all). Patients with AVFS 0 or 1 had restricted mean survival times which were one year longer than those with AVFS 2 or 3 (p < .001), who in turn had restricted mean survival times over one year longer than those with AVFS of 4 or more (p < .001). CONCLUSION: Frailty factors are strong predictors of long term outcomes in vascular surgery. Further prospective studies are warranted to investigate its utility in clinical decision making.