Heterogeneity in Surveillance after Endovascular Aneurysm Repair in the UK

The sac evolution imaging follow-up after endovascular aortic repair: An international expert opinion-based Delphi consensus study

OBJECTIVE

Management of follow-up protocols after endovascular aortic repair (EVAR) varies significantly between centers and is not standardized according to sac regression. By designing an international expert-based Delphi consensus, the study aimed to create recommendations on follow-up after EVAR according to sac evolution.

METHODS

Eight facilitators created appropriate statements regarding the study topic that were voted, using a 4-point Likert scale, by a selected panel of international experts using a three-round modified Delphi consensus process. Based on the experts' responses, only those statements reaching a grade A (full agreement ≥75%) or B (overall agreement ≥80% and full disagreement <5%) were included in the final document.

RESULTS

One-hundred and seventy-four participants were included in the final analysis, and each voted the initial 29 statements related to the definition of sac regression (Q1-Q9), EVAR follow-up (Q10-Q14), and the assessment and role of sac regression during follow-up (Q15-Q29). At the end of the process, 2 statements (6.9%) were rejected, 9 statements (31%) received a grade B consensus strength, and 18 (62.1%) reached a grade A consensus strength. Of 27 final statements, 15 (55.6%) were classified as grade I, whereas 12 (44.4%) were classified as grade II. Experts agreed that sac regression should be considered an important indicator of EVAR success and always be assessed during follow-up after EVAR.

CONCLUSIONS

Based on the elevated strength and high consistency of this international expert-based Delphi consensus, most of the statements might guide the current clinical management of follow-up after EVAR according to the sac regression. Future studies are needed to clarify debated issues.

One-year aneurysm-sac dynamics are associated with reinterventions and rupture following infrarenal endovascular aneurysm repair

OBJECTIVE

One-year aneurysm sac changes have previously been found to be associated with mortality and may have the potential to guide personalized follow-up following endovascular aneurysm repair (EVAR). In this study, we examined the association of these early sac changes with long-term reintervention and rupture.

METHODS

We identified all patients undergoing first-time EVAR for intact abdominal aortic aneurysm between 2003 and 2018 in the Vascular Quality Initiative with linkage to Medicare claims for long-term outcomes. We included patients with an imaging study at 1 year postoperatively. Aneurysm sac behavior was defined as per the Society for Vascular Surgery guidelines: stable sac (<5 mm change), sac regression (≥5 mm), and sac expansion (≥5 mm). Outcomes included mortality, reintervention, and rupture within 8 years, which were assessed with Kaplan-Meier methods and multivariable Cox regression analysis. Secondarily, we utilized polynomial spline interpolation to demonstrate the continuous relationship of diameter change to 8-year hazard of reintervention, rupture, or mortality as a composite outcome.

RESULTS

Of 31,185 EVAR patients, 16,102 (52%) had an imaging study at 1 year and were included in this study. At 1 year, 44% of sacs remained stable, 49% regressed, and 6.2% displayed expansion. Following risk adjustment, compared with a stable sac at 1 year, sac regression was associated with lower 8-year mortality (49% vs 53%; hazard ratio [HR], 0.92; 95% confidence interval [CI], 0.85-0.99; P = .036), reintervention rate (8.9% vs 15%; HR, 0.58; 95% CI, 0.50-0.68; P < .001), and rupture rate (2.0% vs 4.0%; HR, 0.45; 95%CI, 0.29-0.69; P < .001). Conversely, compared with a stable sac, sac expansion was associated with higher 8-year mortality (64% vs 53%; HR, 1.31; 95% CI, 1.14-1.51; P < .001) and reintervention rate (27% vs 15%; HR, 1.98; 95% CI, 1.57-2.51; P < .001), but similar risk of rupture (7.2% vs 4.0%; HR, 1.61; 95% CI, 0.88-2.96; P = .12). Polynomial spline interpolation demonstrated that, compared with no diameter change at 1 year, increased sac regression was associated with an incrementally lower risk of late outcomes, whereas increased sac expansion was associated with an incrementally higher risk of late outcomes.

CONCLUSIONS

Following EVAR, compared with a stable sac at 1-year imaging, sac regression and expansion are associated with a lower and higher risk respectively of long-term mortality, reinterventions, and ruptures. Moreover, the amount of regression or expansion seems to be incrementally associated with these late outcomes, too. Future studies are needed to determine how to improve 1-year sac regression, and whether it is safe to extend follow-up intervals for patients with regressing sacs.

Possible implications of device-specific variability in post-endovascular aneurysm repair sac regression and endoleaks for surveillance categorization

OBJECTIVE

Significant sac regression during early surveillance has been shown to best predict reintervention-free long-term surveillance after endovascular aneurysm repair (EVAR). Furthermore, a persistent endoleak has been related to a worse outcome. Individualized surveillance algorithms based on these findings have been suggested. There are no studies comparing the performance of different stent grafts regarding sac regression, the presence of type II endoleaks, and their possible implications for individualized surveillance. The objective of this study was to evaluate device-specific differences and how these may affect patient categorization for surveillance.

METHODS

Patients were treated electively with standard EVAR between 2005 and 2015 using three different devices (Zenith by Cook, Excluder by Gore, and Endurant by Medtronic). The data were reviewed retrospectively until 2020. Patients' computed tomography angiographies (CTAs) at 30 days and at 2 years were analyzed for freedom from endoleaks and for sac regression of ≥5 mm. Reinterventions during long-term surveillance were counted. Patients were categorized according to the presence of any endoleak and sac regression at 30 days and 2 years, and the probability of reintervention-free long-term surveillance was evaluated based on these findings.

RESULTS

A total of 435 patients were treated for an abdominal aortic aneurysm with EVAR during the study period. At 30 days, 80.0% (n = 339) of the patients were free from endoleaks, and at 2 years, 78.9% (n = 273) were free from endoleaks. There was a significant difference in endoleak rate at 30 days and 2 years between the devices (P < .001 and P = .001). There was no significant difference in sac regression between the devices at 2 years (P = .096). The categorization at 30 days based on endoleak status had a sensitivity of 44.9%, specificity of 87.4%, and negative predictive value of 84.1% for finding a reintervention-requiring complication during long-term follow-up. The corresponding figures at 2 years were 63.3%, 91.4%, and 89.4%, respectively. The combination of freedom from endoleaks and sac regression of ≥5 mm in the 2-year CTA best predicted an uneventful long-term surveillance. Patients who met this criterion had a 95.6% probability (negative predictive value) of having a reintervention-free long-term surveillance.

CONCLUSIONS

There are significant differences in the prevalence of endoleaks between devices at 30 days and 2 years, but there is no difference in sac regression. Patients with sac regression of ≥5 mm and no endoleaks in the 2-year CTA can be safely categorized for infrequent surveillance regardless of the stent graft model that has initially been used.

Postoperative imaging follow-up at 2 years as a predictor of long-term outcomes after endovascular aneurysm repair

OBJECTIVE

Lifelong imaging follow-up remains the reference standard after endovascular aneurysm repair (EVAR). Because the number of EVARs has been increasing, an individually optimized follow-up protocol is desirable. The aim of the present study was to investigate the possibility of creating such a protocol by evaluating patients' follow-up images and to compare the findings with their freedom from reintervention and abdominal aortic aneurysm (AAA) rupture.

METHODS

From 2000 to 2010, 282 patients with an AAA had undergone elective EVAR with the Zenith stent-graft. The patients were followed up annually until the end of 2019. The patients were categorized into two groups according to the presence of any endoleak or sac shrinkage at 2 years. Group A included those with no detectable endoleak plus aneurysm sac shrinkage of ≥5 mm (n = 137; 63%), and group B included those with any type of endoleak and/or no significant aneurysm sac shrinkage (<5 mm; n = 82; 37%).

RESULTS

The mean follow-up was 83 months (range, 0-229 months), and the overall survival at 2 years was 84% (n = 237). No significant difference was found in overall survival between groups A and B (P = .73). However, a significant difference was found in freedom from AAA rupture at 12 years, favoring group A (group A, 100%; group B, 91%; P = .002). Furthermore, in group A, the freedom from reintervention was 95% at 12 years compared with only 31% in group B (P < .001). The sensitivity of the categorization was 89% and the specificity was 74% for the finding of a complication requiring a reintervention during long-term follow-up.

CONCLUSIONS

Patients without an endoleak and a reduction of ≥5 mm in aneurysm size at 2 years had significantly fewer late reinterventions and ruptures during long-term follow-up compared with their counterparts using the Zenith stent-graft (Cook Medical, Bloomington, Ind). This finding suggests that, for these patients, the follow-up interval can be personalized and safely extended after 2 years.

Appropriateness of surgery performed for abdominal aortic aneurysm at tertiary hospitals in Spain

BACKGROUND

To analyze the appropriateness of the type of repair (open or endovascular) performed for abdominal aortic aneurysm (AAA) in five university hospitals in Spain, according to evidence-based recommendations.

METHODS

A multicenter, retrospective cross-sectional study of patients with AAA who underwent elective open surgical repair (OSR) or endovascular aneurysm repair (EVAR). Data were collected on demographic and clinical variables and type of surgical repair. A pair of vascular surgeons from each participating hospital performed a blinded assessment based on GRADE recommendations. The concordance between the two evaluators and the agreement between their evidence-based recommendation and the procedure actually performed were assessed.

RESULTS

A total of 186 patients were selected; 179 were included. Mean age was 72.5 years (standard deviation [SD], 8.4), mean Charlson Comorbidity Index (CCI) was 2.04 (SD, 1.9). OSR was performed in 53.2% (n=99) and EVAR in 46.8% (n=87) of cases. Overall, 65.9% (118/179) of interventions performed were considered appropriate: 50% (47/94) of OSRs and 83.5% (71/85) of EVARs. The patient characteristics were similar for all the hospitals, but the chosen surgical technique did show significant differences among these centers. There were no significant differences among the hospitals in the proportion of cases judged as appropriate, either overall (p=0.346) or for each type of procedure (p=0.531 and p=0.538 for OSR and EVAR, respectively).

CONCLUSIONS

In this study, the majority of the AAA repairs performed were appropriate according to GRADE recommendations. A higher proportion of EVARs were considered appropriate than OSRs. Choice of AAA repair should be standardized through the use of evidence-based clinical practice guidelines, while incorporating patient preferences, to reduce the existing variability and ensure appropriate selection of AAA repair technique.

Editor's Choice - Detection of Late Complications After Endovascular Abdominal Aortic Aneurysm Repair and Implications for Follow up Based on Retrospective Assessment of a Two Centre Cohort

OBJECTIVE

Endovascular aortic aneurysm repair (EVAR) is associated with the risk of late complications and mandates follow up. This retrospective study assessed post-EVAR complications in a two centre cohort. The study evaluated the rate of complications presenting with symptoms vs. those detected by imaging follow up. Additionally, the agreement between DUS and CTA in detecting complications was assessed in patients with both.

METHODS

All EVAR patients from 1998 to 2012 in two centres were included. Complications were classified based on whether they were symptomatic or detected by imaging, as well as based on imaging detection modality (DUS or CTA). For patients who had undergone DUS and CTA within three months of each other, the kappa coefficient of agreement was assessed.

RESULTS

Four hundred and fifty-four patients treated by EVAR were identified. The median follow up time was 5.2 (IQR 2.8-7.6) years. One hundred and eighteen patients (26%) developed 176 complications. One hundred and six (60.2%) of the complications were asymptomatic, and 70 (39.8%) were symptomatic. Two hundred and fifty-three patients had imaging with both modalities within three months of each other; the kappa coefficient for agreement between CTA and DUS for detecting clinically significant complications was 0.91. Regarding CTA as the standard modality, DUS had a sensitivity of 88.8% (95% CI 77.3-95.8%) and a specificity of 99.4% (95% CI 97.1-99.9%). Three of the complications missed by DUS were related to loss of proximal and distal seal, all occurring in patients with short sealing length on first post-operative CT scan.

CONCLUSION

Approximately a quarter of the patients developed complications, the majority of which were asymptomatic, underlining the importance of adequate surveillance. There was good agreement between CTA and DUS in detecting complications. Clinically significant complications related to inadequate seal were missed by DUS, suggesting that CTA still plays an important role in EVAR surveillance.

Long-term radiation exposure in patients undergoing EVAR: Reflecting clinical day-to-day practice to assess realistic radiation burden

Endovascular repair of aortic aneurysms (EVAR) has become an established treatment option currently applied in an increasing numbers of patients with aortic aneurysms. Advantages include reduced surgical trauma, procedural time, intensive care unit and hospital lengths of stay, blood loss as well as morbidity and mortality.The optimal imaging modalities in EVAR follow-up as well as the appropriate intervals between these follow-ups remain subject of controversial discussion. Objective of this study was the evaluation of the realistic radiation exposure and risk estimate postop EVAR treatment.Of the follow-ups required according to the surveillance schedule during the first year post-EVAR, only 68.3% were actually implemented. Of those required from the second year onwards, an average of 70% was actually performed. During the observation period, each patient underwent a mean of 4.3 CTAs. The median ED calculated from all CTAs was 24. 5 mSv. The minimum and maximum cumulative EDs for the entire observation period were 55 mSv and 310 mSv, respectively.

Optimizing Surveillance and Re-intervention Strategy Following Elective Endovascular Repair of Abdominal Aortic Aneurysms

BACKGROUND

EVAR for abdominal aortic aneurysm has an initial survival advantage over OR, but more frequent complications increase costs and long-term aneurysm-related mortality. Randomized controlled trials of EVAR versus OR have shown EVAR is not cost-effective over a patient's lifetime. However, in the EVAR-1 trial, postoperative surveillance may have been sub-optimal, as the importance of sac growth as a predictor of graft failure was overlooked.

METHODS

Real-world data informed a discrete event simulation model of postoperative outcomes following EVAR. Outcomes observed EVAR-1 were compared with those from 5 alternative postoperative surveillance and re-intervention strategies. Key events, quality-adjusted life years and costs were predicted. The impact of using complication and rupture rates from more recent devices, imaging and re-intervention methods was also explored.

RESULTS

Compared with observed EVAR-1 outcomes, modeling full adherence to the EVAR-1 scan protocol reduced abdominal aortic aneurysm (AAA) deaths by 3% and increased elective re-interventions by 44%. European Society re-intervention guidelines provided the most clinically effective strategy, with an 8% reduction in AAA deaths, but a 52% increase in elective re-interventions. The cheapest and most cost-effective strategy used lifetime annual ultrasound in primary care with confirmatory computed tomography if necessary, and reduced AAA-related deaths by 5%. Using contemporary rates for complications and rupture did not alter these conclusions.

CONCLUSIONS

All alternative strategies improved clinical benefits compared with the EVAR-1 trial. Further work is needed regarding the cost and accuracy of primary care ultrasound, and the potential impact of these strategies in the comparison with OR.

Descriptive and follow-up study of patients treated surgically for abdominal aortic aneurysm at tertiary hospitals in Spain

BACKGROUND

The aim of this study was to assess potential variability in the clinical characteristics and treatment of patients undergoing elective surgery for abdominal aortic aneurysm (AAA) across five hospitals in Spain.

METHODS

Multicenter, retrospective cohort study of patients diagnosed with AAA and treated with open surgical repair (OSR) or endovascular aneurysm repair (EVAR). We evaluated clinical and demographic variables, including comorbidity (Charlson Comorbidity Index [CCI]); anatomic characteristics; surgical risk (ASA Score); aneurysm characteristics; and in-hospital and overall mortality. All patients were followed for three years.

RESULTS

A total of 186 patients were included, mean age 72.5 (standard deviation [SD], 8.4), mean CCI 2.04 (SD, 1.9). The surgical technique was EVAR in 46.8% of cases (N.=87) and OSR in 53.2% (N.=99). The in-hospital mortality rate was 2.2%, with no differences between groups. The overall mortality rate during follow-up (mean, 2.9 years) was 24.1% for EVAR versus 8.1% for the OSR group (odds ratio [OR], 3.62; 95% confidence interval [CI], 3.60-3.64; P=0.004). EVAR was the only independent risk factor for mortality (OR, 3.89; 95% CI: 3.87-3.92; P=0.004). Inter-center variability in the type of surgery was high, with EVAR accounting for 19.4% to 75% of the surgical procedures, depending on the treating center (P<0.001).

CONCLUSIONS

In this study the in-hospital mortality rates for elective EVAR and OSR were similar. However, after the follow-up, patients who underwent EVAR had a three-fold greater mortality rate than those treated with OSR. There was substantial inter-hospital variability, underscoring the need to standardize treatment selection in patients who undergo elective surgery for AAA repair.

Contrast-enhanced ultrasound and/or colour duplex ultrasound for surveillance after endovascular abdominal aortic aneurysm repair: a systematic review and economic evaluation

BACKGROUND

Endovascular abdominal aortic aneurysm repair (EVAR) of abdominal aortic aneurysm (AAA) is less invasive than open surgery, but may be associated with important complications. Patients receiving EVAR require long-term surveillance to detect abnormalities and direct treatments. Computed tomography angiography (CTA) has been the most common imaging modality adopted for EVAR surveillance, but it is associated with repeated radiation exposure and the risk of contrast-related nephropathy. Colour duplex ultrasound (CDU) and, more recently, contrast-enhanced ultrasound (CEU) have been suggested as possible, safer, alternatives to CTA.

OBJECTIVES

To assess the clinical effectiveness and cost-effectiveness of imaging strategies, using either CDU or CEU alone or in conjunction with plain radiography, compared with CTA for EVAR surveillance.

DATA SOURCES

Major electronic databases were searched, including MEDLINE, EMBASE, Science Citation Index, Scopus' Articles-in-Press, Cochrane Central Register of Controlled Trials (CENTRAL), Database of Abstracts of Reviews of Effects (DARE) and NHS Economic Evaluation Database from 1996 onwards. We also searched for relevant ongoing studies and conference proceedings. The final searches were undertaken in September 2016.

METHODS

We conducted a systematic review of randomised controlled trials and cohort studies of patients with AAAs who were receiving surveillance using CTA, CDU and CEU with or without plain radiography. Three reviewers were involved in the study selection, data extraction and risk-of-bias assessment. We developed a Markov model based on five surveillance strategies: (1) annual CTA; (2) annual CDU; (3) annual CEU; (4) CDU together with CTA at 1 year, followed by CDU on an annual basis; and (5) CEU together with CTA at 1 year, followed by CEU on an annual basis. All of these strategies also considered plain radiography on an annual basis.

RESULTS

We identified two non-randomised comparative studies and 25 cohort studies of interventions, and nine systematic reviews of diagnostic accuracy. Overall, the proportion of patients who required reintervention ranged from 1.1% (mean follow-up of 24 months) to 23.8% (mean follow-up of 32 months). Reintervention was mainly required for patients with thrombosis and types I-III endoleaks. All-cause mortality ranged from 2.7% (mean follow-up of 24 months) to 42% (mean follow-up of 54.8 months). Aneurysm-related mortality occurred in < 1% of the participants. Strategies based on early and mid-term CTA and/or CDU and long-term CDU surveillance were broadly comparable with those based on a combination of CTA and CDU throughout the follow-up period in terms of clinical complications, reinterventions and mortality. The economic evaluation showed that a CDU-based strategy generated lower expected costs and higher quality-adjusted life-year (QALYs) than a CTA-based strategy and has a 63% probability of being cost-effective at a £30,000 willingness-to-pay-per-QALY threshold. A CEU-based strategy generated more QALYs, but at higher costs, and became cost-effective only for high-risk patient groups.

LIMITATIONS

Most studies were rated as being at a high or moderate risk of bias. No studies compared CDU with CEU. Substantial clinical heterogeneity precluded a formal synthesis of results. The economic model was hindered by a lack of suitable data.

CONCLUSIONS

Current surveillance practice is very heterogeneous. CDU may be a safe and cost-effective alternative to CTA, with CTA being reserved for abnormal/inconclusive CDU cases.

FUTURE WORK

Research is needed to validate the safety of modified, more-targeted surveillance protocols based on the use of CDU and CEU. The role of radiography for surveillance after EVAR requires clarification.

STUDY REGISTRATION

This study is registered as PROSPERO CRD42016036475.

FUNDING

The National Institute for Health Research Health Technology Assessment programme.

Prospective, single UK centre, comparative study of the predictive values of contrast-enhanced ultrasound compared to time-resolved CT angiography in the detection and characterisation of endoleaks in high-risk patients undergoing endovascular aneurysm repair surveillance: a protocol

INTRODUCTION

Diagnosis of endoleaks is imperative to prevent failure of endovascular aneurysm repairs (EVARs). The gold standard for diagnosis of endoleaks is catheter-directed subtraction angiography, which is not a practicable choice for surveillance. CT angiography (CTA) is the historical surveillance modality of choice. Concerns over cost, potential nephrotoxicity of contrast agents and repeated radiation exposure led to colour duplex ultrasound scan (CDUS) becoming an established alternative. CDUS has a lower sensitivity and specificity for endoleaks detection compared to CTA. Contrast-enhanced ultrasound scan (CEUS) represents an improvement of ultrasound imaging but comparisons against CTA report widely varying results, likely due to technical factors of CEUS and limitations of single-phase CTA.The development of time-resolved CTA (tCTA) offers timing information that much more closely mirrors the dynamic information available from CEUS. Theoretically, these two imaging modalities have the best potential for diagnostic accuracy. The aim of this study will be to compare CEUS to tCTA and investigate the utility of other measurements available from tCTA.

METHODS AND ANALYSIS

This is a prospective, single UK centre, comparative study of paired binary diagnostic imaging modalities. Patients identified in routine post-EVAR surveillance as at risk of having a graft-related endoleak will undergo a CEUS and tCTA on the same day. This will allow the first comparison of CEUS to a semidynamic form of CTA. CEUS sensitivity and specificity to endoleak detection will be calculated.

ETHICS AND DISSEMINATION

The study has achieved ethical approval. We hope the results will define the diagnostic accuracy of CEUS in comparison to a semidynamic form of CTA, representing a methodological improvement from previous studies. Results will be submitted for presentation at national and international vascular surgeryandradiology meetings. The full results are planned to be published in a medical journal.

TRIAL REGISTRATION NUMBER

NCT02688751.

Follow-up after endovascular aortic aneurysm repair can be stratified based on first postoperative imaging

Background

Lifelong postoperative surveillance is recommended following endovascular aneurysm repair (EVAR). Although the purpose is to prevent and/or identify complications early, it also results in increased cost and workload. This study was designed to examine whether it may be possible to identify patients at low risk of complications based on their first postoperative CT angiogram (CTA).

Methods

All patients undergoing EVAR in two Swedish centres between 2001 and 2012 were identified retrospectively and categorized based on the first postoperative CTA as at low risk (proximal and distal sealing zone at least 10 mm and no endoleak) or high risk (sealing zone less than 10 mm and/or presence of any endoleak) of complications.

Results

Some 326 patients (273 men) with a CTA performed less than 1 year after EVAR were included (low risk 212, 65·0 per cent; high risk 114, 35·0 per cent). There was no difference between the groups in terms of sex, age, co-morbidities, abdominal aortic aneurysm (AAA) diameter, preoperative AAA neck anatomy, stent-graft type or duration of follow-up (mean(s.d.) 4·8(3·2) years). Five-year freedom from AAA-related adverse events was 97·1 and 47·7 per cent in the low- and high-risk groups respectively (P &lt; 0·001). The corresponding freedom from AAA-related reintervention was 96·2 and 54·1 per cent (P &lt; 0·001). The method had a sensitivity of 88·3 per cent, specificity of 77·0 per cent and negative predictive value of 96·6 per cent to detect AAA-related adverse events. The number of surveillance imaging per AAA-related adverse event was 168 versus 11 for the low-risk versus high-risk group.

Conclusion

Two-thirds of patients undergoing EVAR have an adequate seal and no endoleak on the first postoperative CTA, and a very low risk of AAA-related events up to 5 years. Less vigilant follow-up after EVAR may be considered for these patients.

Identification of Factors Influencing Cumulative Long-Term Radiation Exposure in Patients Undergoing EVAR

Patients who undergo endovascular repair of aortic aneurysms (EVAR) require life-long surveillance because complications including, in particular, endoleaks, aneurysm rupture, and graft dislocation are diagnosed in a certain share of the patient population and may occur at any time after the original procedure. Radiation exposure in patients undergoing EVAR and post-EVAR surveillance has been investigated by previous authors. Arriving at realistic exposure data is essential because radiation doses resulting from CT were shown to be not irrelevant. Efforts directed at identification of factors impacting the level of radiation exposure in both the course of the EVAR procedure and post-EVAR endovascular interventions and CTAs are warranted as potentially modifiable factors may offer opportunities to reduce the radiation. In the light of the risks found to be associated with radiation exposure and considering the findings above, those involved in EVAR and post-EVAR surveillance should aim at optimal dose management.

Using multiple classifiers for predicting the risk of endovascular aortic aneurysm repair re-intervention through hybrid feature selection

Feature selection is essential in medical area; however, its process becomes complicated with the presence of censoring which is the unique character of survival analysis. Most survival feature selection methods are based on Cox's proportional hazard model, though machine learning classifiers are preferred. They are less employed in survival analysis due to censoring which prevents them from directly being used to survival data. Among the few work that employed machine learning classifiers, partial logistic artificial neural network with auto-relevance determination is a well-known method that deals with censoring and perform feature selection for survival data. However, it depends on data replication to handle censoring which leads to unbalanced and biased prediction results especially in highly censored data. Other methods cannot deal with high censoring. Therefore, in this article, a new hybrid feature selection method is proposed which presents a solution to high level censoring. It combines support vector machine, neural network, and K-nearest neighbor classifiers using simple majority voting and a new weighted majority voting method based on survival metric to construct a multiple classifier system. The new hybrid feature selection process uses multiple classifier system as a wrapper method and merges it with iterated feature ranking filter method to further reduce features. Two endovascular aortic repair datasets containing 91% censored patients collected from two centers were used to construct a multicenter study to evaluate the performance of the proposed approach. The results showed the proposed technique outperformed individual classifiers and variable selection methods based on Cox's model such as Akaike and Bayesian information criterions and least absolute shrinkage and selector operator in p values of the log-rank test, sensitivity, and concordance index. This indicates that the proposed classifier is more powerful in correctly predicting the risk of re-intervention enabling doctor in selecting patients' future follow-up plan.

Improving Patient Compliance With Post-EVAR Surveillance May Prevent Late Rupture

Endovascular aneurysm repair (EVAR) has gained increasing popularity in the treatment of infrarenal abdominal aortic aneurysm. Despite its favorable early outcomes, the long-term efficacy of EVAR remains a concern. Late rupture is the ultimate treatment failure and continues to complicate EVAR. Univariate and multivariate analyses have identified factors predictive of late rupture. The importance of EVAR surveillance to prevent late complications is equally widely acknowledged. This article aims to present our current understanding of the precipitating factors of late rupture after EVAR and explores whether the key to its prevention lies within improving patient factors, particularly compliance to follow-up appointments or whether physicians hold the solution.

Feature selection through validation and un-censoring of endovascular repair survival data for predicting the risk of re-intervention

Background

Feature selection (FS) process is essential in the medical area as it reduces the effort and time needed for physicians to measure unnecessary features. Choosing useful variables is a difficult task with the presence of censoring which is the unique characteristic in survival analysis. Most survival FS methods depend on Cox’s proportional hazard model; however, machine learning techniques (MLT) are preferred but not commonly used due to censoring. Techniques that have been proposed to adopt MLT to perform FS with survival data cannot be used with the high level of censoring. The researcher’s previous publications proposed a technique to deal with the high level of censoring. It also used existing FS techniques to reduce dataset dimension. However, in this paper a new FS technique was proposed and combined with feature transformation and the proposed uncensoring approaches to select a reduced set of features and produce a stable predictive model.

Methods

In this paper, a FS technique based on artificial neural network (ANN) MLT is proposed to deal with highly censored Endovascular Aortic Repair (EVAR). Survival data EVAR datasets were collected during 2004 to 2010 from two vascular centers in order to produce a final stable model. They contain almost 91% of censored patients. The proposed approach used a wrapper FS method with ANN to select a reduced subset of features that predict the risk of EVAR re-intervention after 5 years to patients from two different centers located in the United Kingdom, to allow it to be potentially applied to cross-centers predictions. The proposed model is compared with the two popular FS techniques; Akaike and Bayesian information criteria (AIC, BIC) that are used with Cox’s model.

Results

The final model outperforms other methods in distinguishing the high and low risk groups; as they both have concordance index and estimated AUC better than the Cox’s model based on AIC, BIC, Lasso, and SCAD approaches. These models have p-values lower than 0.05, meaning that patients with different risk groups can be separated significantly and those who would need re-intervention can be correctly predicted.

Conclusion

The proposed approach will save time and effort made by physicians to collect unnecessary variables. The final reduced model was able to predict the long-term risk of aortic complications after EVAR. This predictive model can help clinicians decide patients’ future observation plan.

Electronic supplementary material

The online version of this article (doi:10.1186/s12911-017-0508-3) contains supplementary material, which is available to authorized users.

A Systematic Review of Predictors of Reintervention After EVAR: Guidance for Risk-Stratified Surveillance

BACKGROUND

Current surveillance protocols after endovascular aneurysm repair (EVAR) are ineffective and costly. Stratifying surveillance by individual risk of reintervention requires an understanding of the factors involved in developing post-EVAR complications. This systematic review assessed risk factors for reintervention after EVAR and proposals for stratified surveillance.

METHODS

A systematic search according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines was performed using EMBASE and MEDLINE databases to identify studies reporting on risk factors predicting reintervention after EVAR and proposals for stratified surveillance.

RESULTS

Twenty-nine studies reporting on 39 898 patients met the primary inclusion criteria for reporting predictors of reintervention or aortic complications with or without suggestions for stratified surveillance. Five secondary studies described external validation of risk scores for reintervention or aortic complications. There was great heterogeneity in reporting risk factors identified at the pre-EVAR, intraoperative, and post-EVAR stages of treatment, although large preoperative abdominal aortic aneurysm diameter was the most commonly observed risk factor for reintervention after EVAR.

CONCLUSION

Existing data on predictors of post-EVAR complications are generally of poor quality and largely derived from retrospective studies. Few studies describing suggestions for stratified surveillance have been subjected to external validation. There is a need to refine risk prediction for EVAR failure and to conduct prospective comparative studies of personalized surveillance with standard practice.

Endovascular repair of abdominal aortic aneurysms

Endovascular repair of abdominal aortic aneurysms is an important technique in the vascular surgeon's armamentarium, which has created a seismic shift in the management of aortic pathology over the past two decades. In comparison to traditional open repair, the endovascular approach is associated with significantly improved perioperative morbidity and mortality. The early survival benefit of endovascular abdominal aortic aneurysm repair is sustained up to 3 years postoperatively, but longer-term life expectancy remains poor regardless of operative modality. Nonetheless, most abdominal aortic aneurysms are now repaired using endovascular stent grafts. The technology is not perfect as several postoperative complications, namely endoleak, stent-graft migration, and graft limb thrombosis, can develop and therefore lifelong imaging surveillance is required. In addition, a postoperative inflammatory response has been documented after endovascular repair of aortic aneurysms; the clinical significance of this finding has yet to be determined. Subsequently, the safety and applicability of endovascular stent grafts are likely to improve and expand with the introduction of newer-generation devices and with the simplification of fenestrated systems.

An Artificial Neural Network Stratifies the Risks of Reintervention and Mortality after Endovascular Aneurysm Repair; a Retrospective Observational study

Background

Lifelong surveillance after endovascular repair (EVAR) of abdominal aortic aneurysms (AAA) is considered mandatory to detect potentially life-threatening endograft complications. A minority of patients require reintervention but cannot be predictively identified by existing methods. This study aimed to improve the prediction of endograft complications and mortality, through the application of machine-learning techniques.

Methods

Patients undergoing EVAR at 2 centres were studied from 2004-2010. Pre-operative aneurysm morphology was quantified and endograft complications were recorded up to 5 years following surgery. An artificial neural networks (ANN) approach was used to predict whether patients would be at low- or high-risk of endograft complications (aortic/limb) or mortality. Centre 1 data were used for training and centre 2 data for validation. ANN performance was assessed by Kaplan-Meier analysis to compare the incidence of aortic complications, limb complications, and mortality; in patients predicted to be low-risk, versus those predicted to be high-risk.

Results

761 patients aged 75 +/- 7 years underwent EVAR. Mean follow-up was 36+/- 20 months. An ANN was created from morphological features including angulation/length/areas/diameters/volume/tortuosity of the aneurysm neck/sac/iliac segments. ANN models predicted endograft complications and mortality with excellent discrimination between a low-risk and high-risk group. In external validation, the 5-year rates of freedom from aortic complications, limb complications and mortality were 95.9% vs 67.9%; 99.3% vs 92.0%; and 87.9% vs 79.3% respectively (p<0.001)

Conclusion

This study presents ANN models that stratify the 5-year risk of endograft complications or mortality using routinely available pre-operative data.

Patient Compliance with Surveillance Following Elective Endovascular Aneurysm Repair

PURPOSE

Integral to maintaining good outcomes post-endovascular aneurysm repair (EVAR) is a robust surveillance protocol. A significant proportion of patients fail to comply with surveillance, exposing themselves to complications. We examine EVAR surveillance in Wessex (UK), exploring factors that may predict poor compliance.

METHODS

Retrospective analysis of 179 consecutive elective EVAR cases [2008-2013] was performed. 167 patients were male, with the age range of 50-95. Surveillance was conducted centrally (tertiary referral trauma centre) and at four spoke units. Surveillance compliance and predictors of non-compliance including age, gender, co-morbid status, residential location and socioeconomic status were analysed for univariate significance.

RESULTS

Fifty patients (27.9 %) were non-compliant with surveillance; 14 (8.1 %) had no imaging post-EVAR. At 1 year, 56.1 % (of 123 patients) were compliant. At years 2 and 3, 41.5 and 41.2 % (of 65 and 34 patients, respectively) were compliant. Four years post-EVAR, only one of eight attended surveillance (12.5 %). There were no statistically significant differences in age (p = 0.77), co-morbid status or gender (p = 0.64). Distance to central unit (p = 0.67) and surveillance site (p = 0.56) was non-significant. While there was a trend towards compliance in upper-middle-class socioeconomic groups (ABC1 vs. C1C2D), correlating with >50 % of non-compliant patients living within <10 mile radius of the central unit, overall predictive value was not significant (p = 0.82).

CONCLUSIONS

Compliance with surveillance post-EVAR is poor. No independent predictor of non-compliance has been confirmed, but socioeconomic status appears to be relevant. There is a worrying drop-off in attendance beyond the first year. This study highlights a problem that needs to be addressed urgently, if we are to maintain good outcomes post-EVAR.

International validation of a risk score for complications and reinterventions after endovascular aneurysm repair

Background

Lifelong surveillance is considered mandatory after endovascular repair (EVAR) of abdominal aortic aneurysms to detect endograft complications and prevent aneurysm rupture. Current protocols are not cost-effective or clinically effective. The international validity of the St George's Vascular Institute (SGVI) score for EVAR complications was examined.

Methods

The ENGAGE registry recruited patients undergoing EVAR at 79 centres in 30 countries. Reinterventions and endograft complications were recorded for up to 3 years after surgery. Preoperative aneurysm morphology was extracted from the registry database, and used to predict whether patients would be at low or high risk of complications after EVAR based on the SGVI score. Kaplan–Meier analysis was used to compare the incidence of endograft complications and reinterventions in patients predicted to be at low risk compared with those predicted to be at high risk.

Results

Some 1207 patients underwent EVAR, with follow-up of up to 3 years. The SGVI score accurately discriminated freedom from reinterventions (90·5 versus 79·3 per cent in low- versus high-risk patients; P &lt; 0·001), freedom from endograft complications (77·9 versus 69·6 per cent in low- versus high-risk patients; P = 0·012), and freedom from a composite outcome measure of reinterventions or endograft complications (75·0 versus 66·1 per cent in low- versus high-risk patients; P = 0·006) during mid-term follow-up.

Conclusion

This study has provided international validation of a morphological risk score that predicts mid-term reinterventions and endograft complications. The results may enable risk-stratified surveillance after EVAR.

A systematic review and meta-analysis indicates underreporting of renal dysfunction following endovascular aneurysm repair

Deterioration in renal function has been described after endovascular repair of abdominal aortic aneurysms (EVRs). The etiology is multifactorial and represents an important therapeutic target. A need exists to quantitatively summarize incidence and severity of renal dysfunction after EVR to allow better-informed attempts to preserve renal function and improve life expectancy. Here a systematic search was performed using Medline and Embase for renal function after EVR applying PRISMA statements. Univariate and multivariate random-effects meta-analyses were performed to estimate pooled postoperative changes in serum creatinine and creatinine clearance at four time points after EVR. Clinically relevant deterioration in renal function was also estimated at 1 year or more after EVR. Pooled probability of clinically relevant deterioration in renal function at 1 year or more was 18% (95% confidence interval of 14-23%, I2 of 82.5%). Serum creatinine increased after EVR by 0.05 mg/dl at 30 days/1 month, 0.09 mg/dl at 1 month to 1 year, and 0.11 mg/dl at 1 year or more (all significant). Creatinine clearance decreased after EVR by 5.65 ml/min at 1 month-1 year and by 6.58 ml/min at 1 year or more (both significant). Thus, renal dysfunction after EVR is common and merits attention.

CT angiography in the diagnosis of cardiovascular disease: a transformation in cardiovascular CT practice

Computed tomography (CT) angiography represents the most important technical development in CT imaging and it has challenged invasive angiography in the diagnostic evaluation of cardiovascular abnormalities. Over the last decades, technological evolution in CT imaging has enabled CT angiography to become a first-line imaging modality in the diagnosis of cardiovascular disease. This review provides an overview of the diagnostic applications of CT angiography (CTA) in cardiovascular disease, with a focus on selected clinical challenges in some common cardiovascular abnormalities, which include abdominal aortic aneurysm (AAA), aortic dissection, pulmonary embolism (PE) and coronary artery disease. An evidence-based review is conducted to demonstrate how CT angiography has changed our approach in the diagnosis and management of cardiovascular disease. Radiation dose reduction strategies are also discussed to show how CT angiography can be performed in a low-dose protocol in the current clinical practice.

Early sac shrinkage predicts a low risk of late complications after endovascular aortic aneurysm repair

BACKGROUND

Aneurysm shrinkage has been proposed as a marker of successful endovascular aneurysm repair (EVAR). Patients with early postoperative shrinkage may experience fewer subsequent complications, and consequently require less intensive surveillance.

METHODS

Patients undergoing EVAR from 2000 to 2011 at three vascular centres (in 2 countries), who had two imaging examinations (postoperative and after 6-18 months), were included. Maximum diameter, complications and secondary interventions during follow-up were registered. Patients were categorized according to early sac dynamics. The primary endpoint was freedom from late complications. Secondary endpoints were freedom from secondary intervention, postimplant rupture and direct (type I/III) endoleaks.

RESULTS

Some 597 EVARs (71.1 per cent of all EVARs) were included. No shrinkage was observed in 284 patients (47.6 per cent), moderate shrinkage (5-9 mm) in 142 (23.8 per cent) and major shrinkage (at least 10 mm) in 171 patients (28.6 per cent). Four years after the index imaging, the rate of freedom from complications was 84.3 (95 per cent confidence interval 78.7 to 89.8), 88.1 (80.6 to 95.5) and 94.4 (90.1 to 98.7) per cent respectively. No shrinkage was an independent risk factor for late complications compared with major shrinkage (hazard ratio (HR) 3.11; P < 0.001). Moderate compared with major shrinkage (HR 2.10; P = 0.022), early postoperative complications (HR 3.34; P < 0.001) and increasing abdominal aortic aneurysm baseline diameter (HR 1.02; P = 0.001) were also risk factors for late complications. Freedom from secondary interventions and direct endoleaks was greater for patients with major sac shrinkage.

CONCLUSION

Early change in aneurysm sac diameter is a strong predictor of late complications after EVAR. Patients with major sac shrinkage have a very low risk of complications for up to 5 years. This parameter may be used to tailor postoperative surveillance.

Rupture of the infrarenal abdominal aortic aneurysm (AAA) following an endovascular aneurysm repair (EVAR) due to an isolated type II endoleak

We present a case of an elderly man with the persisting type II endoleak following endovascular aneurysm repair. In a view of the multiple comorbidities, the patient was denied an open elective repair. However, when presented with a ruptured aneurismal sac, he underwent an emergency open abdominal aortic aneurysm repair. Overall, the patient has made a good postoperative recovery.

Predicting aortic complications after endovascular aneurysm repair

Background

Lifelong surveillance is standard after endovascular repair of abdominal aortic aneurysm (EVAR), but remains costly, heterogeneous and poorly calibrated. This study aimed to develop and validate a scoring system for aortic complications after EVAR, informing rationalized surveillance.

Methods

Patients undergoing EVAR at two centres were studied from 2004 to 2010. Preoperative morphology was quantified using three-dimensional computed tomography according to a validated protocol, by investigators blinded to outcomes. Proportional hazards modelling was used to identify factors predicting aortic complications at the first centre, and thereby derive a risk score. Sidak tests between risk quartiles dichotomized patients to low- or high-risk groups. Aortic complications were reported by Kaplan–Meier analysis and risk groups were compared by log rank test. External validation was by comparison of aortic complications between risk groups at the second centre.

Results

Some 761 patients, with a median age of 75 (interquartile range 70–80) years, underwent EVAR. Median follow-up was 36 (range 11–94) months. Physiological variables were not associated with aortic complications. A morphological risk score incorporating maximum aneurysm diameter (P &lt; 0·001) and largest common iliac diameter (measured 10 mm from the internal iliac origin; P = 0·004) allocated 75 per cent of patients to a low-risk group, with excellent discrimination between 5-year rates of aortic complication in low- and high-risk groups at both centres (centre 1: 12 versus 31 per cent, P &lt; 0·001; centre 2: 12 versus 45 per cent, P = 0·002).

Conclusion

The risk score uses commonly available morphological data to stratify the rate of complications after EVAR. The proposals for rationalized surveillance could provide clinical and economic benefits.

Surveillance of Endovascular Aneurysm Repair: An Audit of Sonographer-Led Ultrasound and Radiologist-Led Computed Tomography Services

Objectives

Postoperative complications associated with endovascular aneurysm repair (EVAR) mandate the need for lifelong surveillance. The aim of this audit was to compare the detection rates of endoleaks during the first 10 months of a combined sonographer-led colour Doppler ultrasound (CDUS) and radiologist-led contrast-enhanced computed tomography (CECT), postoperative EVAR surveillance programme, within a single National Health Service Hospital.

Methods

A prospective audit was carried out of all patients with EVAR, who had same-day CECT and CDUS surveillance examinations between March 2010 and January 2011. CECT examinations were reported by three consultant vascular radiologists. Of 66 CDUS examinations, 64 (97%) were performed and reported by four certified vascular sonographers. The reports of dual-modality, same-day scans were compared, to establish agreement on the presence and classification of the type, or absence of endoleaks. CECT was used as the gold standard, against which the sensitivity and specificity of CDUS in endoleak detection was determined.

Results

Sixty-six paired same-day CECT and CDUS reports were compared. Ten endoleaks were identified by CECT (15% incidence) and eight (80%) of 10 were type II. The number of observed agreements was 58 (88%) of 66; CDUS missed six endoleaks (five type II) and suggested two false-positives. CDUS had a sensitivity of 40% and a specificity of 96%. Conclusions: In our study, CDUS demonstrated an excellent specificity but had a low sensitivity for endoleak detection. We agree CDUS cannot effectively replace CECT as the sole imaging modality. Research to determine an international consensus for an effective and efficient postoperative EVAR surveillance patient pathway is required.

Systematic review and meta-analysis of duplex ultrasonography, contrast-enhanced ultrasonography or computed tomography for surveillance after endovascular aneurysm repair

BACKGROUND

Previous analyses suggested that duplex ultrasonography (DUS) detected endoleaks after endovascular aneurysm repair (EVAR) with insufficient sensitivity; they did not specifically examine types 1 and 3 endoleak, which, if untreated, may lead to aneurysm-related death. In light of changes to clinical practice, the diagnostic accuracy of DUS and contrast-enhanced ultrasonography (CEUS) for types 1 and 3 endoleak required focused reappraisal.

METHODS

Studies comparing DUS or CEUS with computed tomography (CT) for endoleak detection were identified. CT was taken as the standard in bivariable meta-analysis.

RESULTS

Twenty-five studies (3975 paired scans) compared DUS with CT for all endoleaks. The pooled sensitivity was 0·74 (95 per cent confidence interval 0·62 to 0·83) and the pooled specificity was 0·94 (0·90 to 0·97). Thirteen studies (2650 paired scans) reported detection of types 1 and 3 endoleak by DUS; the pooled sensitivity of DUS was 0·83 (0·40 to 0·97) and the pooled specificity was 1·00 (0·97 to 1·00). Eleven studies (961 paired scans) compared CEUS with CT for all endoleaks. The pooled sensitivity of CEUS was 0·96 (0·85 to 0·99) and the pooled specificity was 0·85 (0·76 to 0·92). Eight studies (887 paired scans) reported detection of types 1 and 3 endoleak by CEUS. The pooled sensitivity of CEUS was 0·99 (0·25 to 1·00) and the pooled specificity was 1·00 (0·98 to 1·00).

CONCLUSION

Both CEUS and DUS were specific for detection of types 1 and 3 endoleak. Estimates of their sensitivity were uncertain but there was no evidence of a clinically important difference. DUS detects types 1 and 3 endoleak with sufficient accuracy for surveillance after EVAR.

Aortic rupture and sac expansion after endovascular repair of abdominal aortic aneurysm

Background

Long-term concerns about the durability of endovascular aortic aneurysm repair (EVAR) remain after the publication of controlled trials. Increased expertise in endograft technology, case selection and postoperative reintervention has created a need for reappraisal of the longer-term efficacy of EVAR using contemporary data.

Methods

Patients undergoing infrarenal EVAR between 2004 and 2010 were studied prospectively. Morphological compliance with manufacturers' instructions for use (IFU) was established using three-dimensional computed tomography. The primary outcome measures were all-cause and aneurysm-related mortality, postoperative rupture, reintervention and sac expansion. These adverse events were reported using Kaplan–Meier survival analysis, with comparison within, or outside IFU by the log rank test.

Results

Some 478 patients of median age 76 years had a median aneurysm diameter of 62·9 mm. Median follow-up was 44 (range 11–94) months; 198 (41·4 per cent) were compliant with IFU. The 30-day mortality rate was 2·1 per cent (10 of 478 patients): nine (2·0 per cent) of 455 patients who had elective and one (4 per cent) of 23 patients who had non-elective surgery. Aneurysm-related mortality was 0·897 deaths per 100 person-years, and all-cause mortality was 8·558 deaths per 100 person-years, with significantly lower survival outside IFU (P = 0·012). Two patients had a late rupture (0·138 per 100 person-years), of whom one died. There were 6·120 reinterventions per 100 person-years, with no difference for aneurysms treated outside IFU (P = 0·136). Primary sac expansion occurred in 6·721 per 100 person-years and secondary sac expansion in 4·142 per 100 person-years.

Conclusion

In this series EVAR had a lower aneurysm-related mortality rate than demonstrated in early controlled trials, and with lower sac expansion rates than reported from image repositories. Data from earlier studies should be applied to current practice with caution.

Operator-controlled imaging significantly reduces radiation exposure during EVAR

INTRODUCTION

Adoption of endovascular aneurysm repair (EVAR) has led to significant reductions in the short-term morbidity and mortality associated with abdominal aortic aneurysm (AAA) repair. However, EVAR may expose both patient and interventionalist to potentially harmful levels of radiation, particularly as more complex procedures are undertaken. The aim of this study was to assess whether changing from radiographer-controlled imaging to a system of operator-controlled imaging (OCI) would influence radiation exposure, screening time or contrast dose during EVAR.

METHOD

Retrospective analysis identified patients that had undergone elective EVAR for infra-renal AAA before or after the change to operator-controlled imaging. Data were collected for radiation dose (measured as dose area product; DAP), screening time, total delivered contrast volume and operative duration. Data were also collected for maximum aneurysm diameter, patient age, gender and body mass index.

RESULTS

122 patients underwent EVAR for infra-renal AAA at a single centre between January 2011 and December 2011. 57 of these were prior to installation of OCI and 65 after installation. Median DAP was significantly lower after installation of OCI (4.9 mGy m(2); range 1.25-13.3) than it had been before installation (6.9 mGy m(2); range 1.91-95.0) (p = 0.005). Median screening times before and after installation of OCI were 20.0 min and 16.2 min respectively (p = 0.027) and median contrast volumes before and after the change to OCI were 100 ml and 90 ml respectively (p = 0.21).

CONCLUSION

Introduction of operator-controlled imaging can significantly reduce radiation exposure during EVAR, with particular reduction in the number of 'higher-dose' cases.