If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Editor's Choice – Post-operative Surveillance and Long Term Outcome after Endovascular Aortic Aneurysm Repair in Patients with an Initial Post-operative Computed Tomography Angiogram Without Abnormalities: the Multicentre Retrospective ODYSSEUS Study
Corresponding author. Amsterdam University Medical Centres, University of Amsterdam, Amsterdam Cardiovascular Sciences, Department of Surgery, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands.
Lifelong imaging surveillance is recommended following endovascular aneurysm repair (EVAR). This study aimed to examine the association between adherence to post-operative surveillance and survival and secondary interventions in patients with an initial post-operative computed tomography angiogram (CTA) without abnormalities.
Methods
All consecutive patients undergoing EVAR for intact abdominal aortic aneurysm (AAA) in 16 hospitals between 2007 and 2012 were identified retrospectively, with follow up until December 2018. Patients were included if the initial post-operative CTA showed no types I – III endoleak, kinking, infection, or limb occlusion. Discontinued follow up was defined as at least one 16 month period in which no imaging surveillance was performed. Primary outcomes were aneurysm related mortality and secondary interventions, and secondary outcome all cause mortality. Kaplan–Meier analysis was used to estimate survival, and Cox regression analyses to identify the association between independent variables and outcome. Sensitivity analyses were performed by varying the definition of continued yearly follow up. The study protocol was published (bmjopen-2019-033584).
Results
1 596 patients (552 continued, 1 044 discontinued follow up) were included with a median (interquartile range) follow up of 89.1 months (52.6). Cumulative aneurysm related, overall, and intervention free survival was 99.4/94.8/96.1%, 98.5/72.9/85.9%, and 96.3/45.4/71.1% at 1, 5, and 10 years, respectively. American Society of Anesthesiologists (ASA) classification (ASA IV hazard ratio [HR] 3.810, 95% confidence interval [CI] 1.296 – 11.198), increase in AAA diameter (HR 3.299, 95% CI 1.408 – 7.729), and continued follow up (HR 3.611, 95% CI 1.780 – 7.323) were independently associated with aneurysm related mortality. The same variables and age (HR 1.063 per year, 95% CI 1.052 – 1.074) were significantly associated with all cause mortality. No difference in secondary interventions was observed between patients with continued vs. discontinued follow up (89/552; 16% vs. 136/1044; 13%; p = .091). Sensitivity analyses showed worse aneurysm related and overall survival in patients with continued follow up.
Conclusion
Discontinued follow up is not associated with poor outcomes. Future prospective studies are indicated to determine in which patients imaging follow up can be safely reduced.
Endovascular aneurysm repair (EVAR) has become the predominant modality for the treatment of infrarenal abdominal aortic aneurysms (AAAs) in The Netherlands. There are concerns regarding the long term durability of EVAR, lifelong follow up routines, and adherence to these surveillance programmes. This study has highlighted that discontinued imaging surveillance post EVAR is common in The Netherlands. Moreover, discontinued yearly follow up in patients with initial post-operative computed tomography angiography without abnormalities is not associated with poor outcomes (in terms of mortality or secondary interventions), suggesting that less frequent surveillance may be envisaged. Future prospective studies are indicated to determine in which patient groups follow up can be safely reduced.
Introduction
In current practice, endovascular aneurysm repair (EVAR) is the preferred surgical treatment of asymptomatic infrarenal abdominal aortic aneurysms (AAAs). In The Netherlands, 79% of elective AAA repairs are performed using EVAR.
However, there is a difference in durability between EVAR and OSR. Long term results of previous clinical trials reveal equal overall survival rates, but significantly more secondary interventions in patients treated by EVAR.
Endovascular versus open repair of abdominal aortic aneurysm in 15-years’ follow-up of the UK endovascular aneurysm repair trial 1 (EVAR trial 1): a randomised controlled trial.
Due to a continued risk of endograft related complications and rupture following EVAR, strict imaging surveillance is considered mandatory for all patients, and recommended by current international guidelines
Previous research has not demonstrated any beneficial effect on survival when comparing patients with continued and discontinued imaging surveillance; however, increased secondary intervention rates were observed in patients with continued follow up.
The initial post-operative computed tomography angiogram (CTA) guides patient stratification during follow up. This scan either shows complications, such as endoleaks, occlusion, or endograft malposition, or the absence thereof. If the initial CTA shows no abnormalities the risk of a secondary intervention is low.
Adequate seal and no endoleak on the first computed tomography angiography as criteria for no additional imaging up to 5 years after endovascular aneurysm repair.
Doubts remain as to whether yearly imaging surveillance is essential for all EVAR patients, and if patients whose initial post-operative CTA showed no abnormalities could be considered for less intensive follow up, as suggested by the European Society for Vascular Surgery (ESVS) guidelines.
It is hypothesised that there would be no difference in aneurysm related mortality between patients with continued vs. discontinued follow up, and that patients with continued imaging surveillance undergo more secondary interventions.
Materials and methods
This multicentre retrospective observational study was performed in accordance with the STROBE guidelines,
and complied with the Declaration of Helsinki. The study was approved by the local Institutional Review Board (W18_102#18.130), registered with The Netherlands Trial Registration (NL6953), and the protocol was published.
Postoperative surveillance and long-term outcome after endovascular aortic aneurysm repair in the Netherlands: study protocol for the retrospective ODYSSEUS study.
In accordance with the Dutch Code of Civil Procedure, the opt out procedure was used; by default patient records were used for research, unless the patient objected to participation within four weeks of notification.
Study population
All consecutive patients at 16 academic and teaching hospitals in The Netherlands (one hospital did not participate) with an AAA who underwent elective EVAR between January 2007 and January 2012 were included. This selection provided a maximum follow up of 6 – 11 years on 1 December 2018. Inclusion criteria were the presence of an intact infrarenal aortic or aorto-iliac aneurysm treated by standard EVAR, and a CTA without abnormalities (no type I – III endoleak, kinking, infection or limb occlusion) performed within 90 days of the initial procedure. Patients with ruptured or isolated iliac aneurysms or requiring chimney or fenestrated EVAR were excluded.
Data collection
Data were collected from patient medical records by two investigators (A.G., S.M.) and entered into an electronic database built with Castor EDC (ISO 27001 and NEN 7510 certified).
Baseline characteristics included age, sex, American Society of Anesthesiologists (ASA) classification and endograft type. Anatomical characteristics included pre-operative AAA diameter, neck length, neck angulation, and iliac artery diameters. Follow up data comprised all imaging studies (duplex ultrasonography [DUS], CTA, magnetic resonance angiography) including the AAA diameter, abnormal findings, interventions, and death. Also, imaging studies made for other reasons outside protocol were used. Data were censored at the date when either the patient died or when the end of follow up on 1 December 2018 was reached. Deaths were ascertained by linking the records of the study population and the National Death Register based on sex, name, and date of birth.
Surveillance programmes
All participating centres provided their respective surveillance protocol (Table 1). These local surveillance protocols dated from 2018 and earlier. Follow up protocols varied per institution, in the first year CTA and DUS were most frequently used and usually DUS annually thereafter.
Table 1Surveillance programmes after endovascular repair of an intact abdominal aortic aneurysm in participating centres in The Netherlands
Centre
First imaging
Yearly thereafter
1
A CT angiogram (CTA) after 4–6 w and clinical visit. No abnormalities: yearly duplex ultrasound (DUS). Abnormalities: CTA is repeated. Endoleak type I/III: intervention
DUS
2
Two weeks clinical visit. A CTA, plain film Xray and clinical review at3 mo post-procedure. CTA at 12 mo. DUS + plain film Xray at year 2, 3, and 5. CTA at 5 y post procedure
DUS, 1/5 y CTA
3
6 week CTA. No abnormalities: DUS after 6 mo and thereafter annually. Type I endoleak: intervention or strict surveillance, type II: normal follow up
DUS, CTA if new endoleak and/or sac size increase
4
First day after discharge consult via phone. Clinical visit after two weeks. Six weeks clinical review + CTA
DUS
5
CTA after 4 w
DUS + ABI + biplanar plain film Xray. CTA if new endoleak and/or sac size increase
6
Clinical visit + DUS 6–8 w. DUS after 3 mo.
DUS + plain film Xray every 2 y
7
CTA at 6 w, colour flow DUS at 6 mo
Colour flow DUS On indication: CTA
8
CTA at 4 w and 12 mo
DUS
9
CTA at 4 w and 12 mo
DUS
10
CTA at 4 w and 12 mo
DUS, 1/5 y CTA
11
6 w clinical visit + CTA
DUS or CTA every 5 years
12
CTA at 6 w, CTA or DUS + plain film Xray at 6 mo
DUS + plain film Xray
13
6 w clinical visit + DUS, DUS at 3 mo
DUS
14
CTA at 4 w and 12 mo. No abnormalities: yearly DUS. Abnormalities: CTA at 3 mo, endoleak I/III: intervention
DUS
15
CTA at 6 w + biplanar plain film Xray. DUS at 6 mo
DUS, CTA if sac size increase
16
Pre-discharge: biplanar plain film Xray. CTA at 6 weeks, CTA at 12 months + biplanar plain film Xray
DUS, CTA if new endoleak and/or sac size increase
Abnormalities first CTA:
•
Type I endoleak: intervention
•
Type II endoleak: strict follow up protocol
•
After 6 mo: biplanar plain film Xray + CT, if sac size increase >10% → CTA
•
After 12 mo: biplanar plain film Xray + CT
No abnormalities first CTA:
•
Minimal follow up
•
After 12 mo: biplanar plain film Xray + CT
CT = computed tomography; CTA = computed tomography angiography; DUS = duplex ultrasonography; ABI = ankle brachial index.
Continued follow up was defined as undergoing imaging surveillance at least once every 16 months throughout the entire follow up period, as most patients were rescheduled within three to four months if they missed their annual follow up visit (consensus decision of the study group). Patients missing at least one follow up visit (imperfect follow up) were included as discontinued to imaging follow up. The reasons why patients missed their follow up visit could not be retrieved from their medical files. Imaging studies were evaluated by radiologists from participating centres and not re-analysed for the purpose of this study. Abnormal radiological findings during follow up were recorded as follows: type I – IV endoleak, endograft migration > 10 mm, endograft infection, endograft kinking, limb occlusion, or sac growth (anteroposterior [AP] diameter).
Patients were distributed over three groups depending on the reported aneurysm diameter over time: increased, stable, or decreased. Sac growth and shrinkage were determined as an increase or decrease of > 5 mm between two consecutive imaging studies, or if there was > 5 mm growth or shrinkage in comparison with the initial post-operative AAA diameter. The aneurysm diameter was recorded as stable if less than 5 mm sac growth or shrinkage occurred.
The primary outcome parameters of this study were aneurysm related death and secondary interventions. Secondary outcomes included all cause mortality, radiological findings during follow up and aneurysm rupture.
Statistical analysis
The sample size calculation was based on a superiority design for secondary interventions and non-inferiority for aneurysm related mortality.
Postoperative surveillance and long-term outcome after endovascular aortic aneurysm repair in the Netherlands: study protocol for the retrospective ODYSSEUS study.
In patients with a post-operative CTA without abnormalities and to correct for 10% incomplete or missing data, a total of 1 598 (1 438/0.9) patients were required to detect a 7% difference for the outcome secondary interventions, and 1 451 (1 306/0.9) patients for aneurysm related death with a non-inferiority limit of 3%.
Postoperative surveillance and long-term outcome after endovascular aortic aneurysm repair in the Netherlands: study protocol for the retrospective ODYSSEUS study.
The Shapiro–Wilk test, Q-Q plots, and histograms were used to determine whether continuous data were normally distributed. Continuous variables were presented as mean and standard deviation (SD) for normally distributed data, or median and interquartile range (IQR) for non-normally distributed data. Differences between groups were assessed using Student’s t test or Mann–Whitney U test, as appropriate. Categorical variables were presented as counts and percentages, differences between groups were analysed using the Pearson χ2 test or Fisher’s exact test. The Kaplan–Meier method was used to estimate survival, and the log rank test to compare survival estimates between groups. Time to aneurysm related and all cause mortality was calculated between start of follow up and time of death or December 2018. Secondary interventions were censored at the last contact with the hospital for imaging. To account for missing data, ten imputed datasets were constructed using multiple imputation by predictive mean matching and pooled with Rubin’s rule. Cox regression analyses controlling for ASA classification, age, sex, endograft type, initial AAA diameter, neck length, angulation, maximum iliac diameter, and change in AAA diameter were used to determine the hazard ratio (HR) and 95% confidence interval (CI) of the variables associated with death and secondary interventions. Continued follow up was investigated in Cox regression analyses to adjust the effect of continued imaging surveillance by possible confounders. To correct for misclassification and because of progressive insight, sensitivity analyses were performed by varying the definition of continued follow up. In this first sensitivity analysis, all patients were reclassified and allocated as continued with imaging surveillance if they underwent at least 80% of their required follow up visits. In the second sensitivity analysis continued follow up rates in the participating centres were divided into low, medium or high continued follow up rates to evaluate strategies instead of patients. With these new variables (continued follow up 80% and continued follow up participating centres) Cox regression analyses were repeated. The level for statistical significance was set at a two sided p < .050. All statistical analyses were calculated with SPSS software version 26 and RStudio 3.6.1.
Results
In total, 1 596 patients were included (Supplementary Figure S1), 552 patients had continued yearly imaging surveillance according to the definition, and 1 044 patients discontinued. The mean (SD) age was 73.5 years (7.8) and there was preponderance of males (n = 1 425; 89%). Patient characteristics are presented in Table 2. The amount of imputed missing data ranged from 0.1% to 37.5%. Neck angulation was missing from > 50% of the radiology reports, and therefore excluded from further analyses. Adherence to yearly imaging surveillance varied widely between participating centres from 15.8% to 54.7% (Supplementary Table S1). The percentage of follow up visits declined in later years following primary EVAR (Supplementary Table S2). The 30 day mortality rate following primary EVAR was 0.4% (7/1 596).
Table 2Baseline characteristics in 1 596 patients with continued vs. discontinued follow up after endovascular aneurysm repair of an intact abdominal aortic aneurysm (AAA)
Variable
Missing
Continued follow up (n = 552)
Discontinued follow up (n = 1 044)
p value
Age – y
0
73.4 ± 0.314
73.6 ± 0.248
.52
Male sex
0
495 (89.7)
930 (89.2)
.86
AAA diameter – cm
2
6.1 ± 0.041
6.0 ± 0.030
.39
ASA classification
0
.91
I
6 (1.1)
7 (0.7)
II
225 (40.9)
424 (40.7)
III
283 (51.5)
543 (52.1)
IV
36 (6.5)
68 (6.5)
Unknown
2 (0.4)
2 (0.2)
Endograft
0
.032
Endurant (Medtronic)
240 (43.5)
383 (36.7)
Zenith (Cook)
175 (31.7)
381 (36.5)
Excluder (Gore)
70 (12.7)
115 (11.0)
Talent (Medtronic)
22 (4.0)
74 (7.1)
Powerlink (Endologix)
2 (0.4)
6 (0.6)
Anaconda (Vascutek)
4 (0.7)
2 (0.2)
Other/Unknown
39 (7.0)
83 (7.9)
Neck length – cm
468
3.1 ± 0.062
3.0 ± 0.041
.053
Neck angulation – degrees
1 336
48.3 ± 2.95
54.8 ± 2.87
.12
Maximum iliac diameter – cm
598
1.9 ± 0.060
1.9 ± 0.041
.46
Data are presented as n (%) or mean ± standard error of the mean. ASA = American Society of Anesthesiologists.
As of 1 December 2018, there were 807 deaths of 1 596 EVAR patients: 320 patients died of non-aneurysm related causes, 34 patients of aneurysm related causes (20 with continued and 14 with discontinued follow up), and in 453 patients the cause of death was unknown. The median (IQR) follow up until death or December 2018 for the entire cohort was 89.1 months (52.6). The median (IQR) follow up for patients with continued follow up was 75.4 months (66.1), while for patients with discontinued follow up it was 92.4 months (41.4), p < .001. Freedom from aneurysm related death was 96.9% and 92.0% at five and 10 years for patients with continued follow up, and 99.7% and 98.0% at five and 10 years for patients with discontinued follow up (log rank; p < .001, Fig. 1).
Figure 1Cumulative Kaplan–Meier estimate of aneurysm related mortality in 1 596 patients with continued and discontinued imaging surveillance after endovascular aneurysm repair (EVAR) of an intact abdominal aortic aneurysm.
There was no difference between the two groups in the number of patients undergoing secondary interventions (89/552; 16% vs. 136/1 044; 13%; p = .091). Also, no difference was observed in numbers of acute secondary interventions (symptomatic presentation) between patients with continued and discontinued follow up (16/89; 18% vs. 30/136; 22%; p = .48). The median (IQR) time to a secondary intervention was 56 months (61.7). The most common indications were type I endoleak and limb occlusion (Table 3). The median (IQR) follow up for the entire cohort until last contact with the hospital was 65 months (61.0), and no difference in follow up duration was observed between patients with continued or discontinued imaging follow up (p = .33). Freedom from secondary interventions was higher in patients with discontinued follow up (Log rank; p = .003) (Fig. 2). Cumulative intervention free survival was 96.1%, 85.9%, and 71.1% at one, five, and 10 years, respectively.
Table 3Complications and secondary interventions within five years and in total in 552 and 1 044 patients with continued vs. discontinued follow up following endovascular aneurysm repair of an intact abdominal aortic aneurysm
Radiological finding
Interventions in patients with continued follow up
Interventions in patients with discontinued follow up
Figure 2Cumulative Kaplan–Meier estimate of intervention free survival in patients with continued and discontinued imaging surveillance after endovascular aneurysm repair (EVAR) of an intact abdominal aortic aneurysm.
Overall survival was higher in patients with discontinued imaging follow up (log rank; p < .001) (Supplementary Figure S2). The overall survival rates were 94.8%, 72.9%, and 45.4% at one, five, and 10 years, respectively. Overall survival was 80.3% and 49.6% at five and 10 years for patients with discontinued imaging follow up, and 58.6% and 35.5% at five and 10 years for patients with continued follow up.
When comparing patients with continued vs. discontinued follow up, abnormal radiological findings were seen more often in the “continued follow up” group (192/552; 35% vs. 226/1 044; 22%, p < .001) (Table 4). On imaging many patients had multiple abnormal findings, so this resulted in 739 abnormal findings during follow up. The commonest abnormal radiological findings in patients with continued follow up were sac growth and type II endoleak.
Table 4Number of abnormal findings during surveillance imaging linked to adherence to imaging follow up in 552 and 1 044 patients with continued vs. discontinued follow up following endovascular aneurysm repair of an intact abdominal aortic aneurysm
Twenty-nine patients (1.8%) presented with aneurysm rupture, equally divided between the groups (10/552; 1.8% vs. 19/1044; 1.8%). The causes of aneurysm rupture were: type I endoleak in 15 patients (3/15 previously detected on imaging surveillance), type II endoleak in three patients (2/3 previously detected), one patient with type III endoleak (not previously detected), two patients with endograft kinking (not previously detected), and in eight patients no cause or previously detected abnormalities were recorded. The median (IQR) time to rupture was 73 months (44.7). The median diameter (IQR) at the time of rupture was 7.8 cm (1.9). The majority of ruptures (18/29; 62%) occurred more than five years after the primary procedure. The cumulative freedom from AAA rupture was 98.9% and 94.5% at five and 10 years, respectively.
Cox regression analyses
Table 5 gives results of the Cox regression analyses for both primary outcomes. In multivariable analysis, ASA classification (ASA IV HR 3.810, 95% CI 1.296 – 11.198), change in AAA diameter (increase HR 3.299 per cm, 95% CI 1.408 – 7.729, decrease HR 0.256 per cm, 95% CI 0.096 – 0.688) and continued follow up (HR 3.588, 95% CI 1.760-7.314) were significantly related to aneurysm related death.
Table 5Cox regression univariable and multivariable analyses of the effect of different variables on aneurysm related mortality and secondary interventions in 1 596 patients following endovascular aneurysm repair of an intact abdominal aortic aneurysm (AAA)
In multivariable analysis pre-operative AAA diameter (HR 1.255 per cm, 95% CI 1.108 – 1.422), maximum iliac diameter (HR 1.171 per cm, 95% CI 1.018 – 1.347), increase in AAA diameter (HR 7.335 per cm, 95% CI 4.608 – 11.677), age (HR 0.962 per year, 95% CI 0.945 – 0.978), endograft type (Gore Excluder HR 0.558, 95% CI 0.326 – 0.958), and continued follow up (HR 1.368, 95% CI 1.021-1.831) were significantly associated with secondary interventions.
In univariable analysis of all cause mortality age, AAA diameter, ASA classification, endograft type, maximum iliac diameter, change in AAA diameter, and continued follow up were statistically significant. In multivariable analysis, age (HR 1.063 per year, 95% CI 1.052 – 1.074), ASA classification (ASA IV HR 1.498, 95% CI 1.136 – 1.975), change in AAA diameter (increase HR 0.495 per cm, 95% CI 0.385 – 0.637, decrease HR 0.418 per cm, 95% CI 0.357– 0.489), and continued follow up (HR 1.957, 95% CI 1.688 – 2.268) were significantly associated with all cause mortality. All analyses before imputation are displayed in Supplementary Table S3.
Sensitivity analyses
After applying the alternative definition of continued follow up in which patients were classified as continued imaging surveillance if they had 80% of their follow up visits, 1 018 patients had continued follow up and 578 discontinued. In multivariable analysis, maximum iliac diameter (HR 1.370 per cm, 95% CI 1.030 – 1.821), decrease in AAA diameter (HR 0.299, 95% CI 0.095 – 0.943), and continued follow up, 80% (HR 4.706, 95% CI 1.371 – 16.152) were significantly associated with aneurysm related mortality. In multivariable analysis AAA diameter (HR 1.346 per cm, 95% CI 1.164 – 1.556), maximum iliac diameter (HR 1.316 per cm, 95% CI 1.123 – 1.543), increase in AAA diameter (HR 7.161, 95% CI 3.871 – 13.247), age (HR 0.967, 95% CI 0.946 – 0.989), and continued follow up, 80% (HR 1.774, 95% CI 1.157 – 2.719) were significantly associated with secondary interventions. Multivariable analysis showed that age (HR 1.067, 95% CI 1.055 – 1.078), ASA classification (ASA IV HR 1.511, 95% CI 1.144 – 1.995), change in AAA diameter (increase HR 0.440 per cm, 95% CI 0.341 – 0.567, decrease HR 0.392 per cm, 95% CI 0.334 – 0.459), and continued follow up 80% (HR 1.932, 95% CI 1.651 – 2.260) were significantly associated with all cause mortality (Supplementary Table S3). There was a significant difference between the two groups in the proportion of patients undergoing secondary interventions (168/1 018; 17% vs. 57/578; 10%; p < .001).
Continued imaging follow up in participating centres was classified as low (up to 26%), medium (up to 36%) or high (up to 55%). Multivariable analyses showed no significant association between this distribution of continued follow up rates in participating centres and all outcomes (aneurysm related mortality, secondary interventions, and all cause mortality) (Supplementary Table S3).
Discussion
The current study has endeavoured to address the evidence gap regarding the value of imaging surveillance in post-EVAR patients whose initial post-operative CTA shows no abnormalities. All participating hospitals used their own surveillance protocol with many similarities to the ESVS guidelines.
Most patients who underwent EVAR did not undergo yearly imaging. There was no difference in the number of secondary interventions between patients with continued and discontinued imaging follow up. However, secondary interventions were seen earlier in the group with continued post-EVAR surveillance. This was confirmed after correcting for confounders by multivariable Cox regression. Finally, in the univariable analysis it was found that continued imaging surveillance was associated with decreased survival, an outcome that remained consistent after adjusting for confounders.
The value of the initial post-operative imaging has been investigated before, demonstrating the risk of complications and interventions is low if no abnormalities are shown on initial imaging.
Adequate seal and no endoleak on the first computed tomography angiography as criteria for no additional imaging up to 5 years after endovascular aneurysm repair.
In this study, patients with continued imaging surveillance underwent 3% more secondary interventions. This is less than the expected difference of 7%, which may be explained by the fact that physicians have become more restrained in intervening nowadays. These findings suggest that not all complications require secondary intervention but may be monitored. It is not known whether if abnormalities are seen on imaging, patients tend to undergo stricter follow up.
However, in sensitivity analysis, including the continued follow up 80% variable, patients with continued imaging surveillance underwent more secondary interventions. This finding was similar to Garg et al.,
who found more interventions (10% vs. 1.4%) in patients with complete against incomplete surveillance. The intervention free survival rates were in line with previous studies.
The results showed poor survival outcomes in patients with continued imaging surveillance. Overall survival was 49.6% at 10 years for patients with discontinued follow up vs. 35.5% for patients with continued follow up. Aneurysm related survival was also worse: 98.0% and 92.0% for patients with discontinued and continued follow up, respectively. This is an intriguing fact that needs to be looked at carefully and which, considering the non-inferiority study design, was unexpected. Multiple explanations for this finding may be proposed. First, patients with continued follow up could have multiple comorbidities and have imaging studies for other reasons, and therefore attend the hospital more often. Second, symptomatic patients and patients with abnormalities seen on imaging surveillance tend to adhere better to their surveillance programme. This finding was robust in the sensitivity analysis in which similar variables were independently associated with all cause mortality. However, other factors were associated with aneurysm related mortality which is probably explained by the low incidence of recorded aneurysm related deaths. High rates of continued follow up in participating centres did not lead to better outcomes. Current guidelines focus on assessment of AAA diameter with DUS and recommend CTA if the diameter increases. Contrary to what might be expected, it was found that an increase in AAA diameter was associated with lower all cause mortality. There is no explanation for this finding, which might be due to other comorbidities that were not included in the dataset (residual confounding). During follow up, some patients had several abnormal radiological findings, or even findings that usually require secondary interventions. Yet these scenarios failed to encourage adherence to surveillance protocols, as only 50% of patients underwent yearly imaging surveillance in the first five years following primary EVAR.
or the physician had indicated that the imaging frequency could be reduced in patients too frail to undergo a secondary intervention.
The study showed that 73.0% of patients survived more than five years after EVAR, while 45.6% survived more than 10 years after EVAR, equivalent to another long term study.
As life expectancy is rising, long term endograft durability is of major importance. Based on the present study, the durability of the implanted devices is satisfactory, as the intervention free survival was 71.1% at 10 years. However, newer generation stent grafts will need strict surveillance as their long term results are still unknown, and new technology failures must be avoided as evidenced by a device that was designed to provide ”endovascular aneurysm sealing’’.
This study has limitations including its retrospective observational study design with a risk of information bias. It was unclear if patients were lost to follow up, if imaging studies had been discontinued after uncomplicated surveillance, or if patients had undergone imaging surveillance elsewhere. Nonetheless, the long term results of a treatment method also require retrospective analysis, since not all potential problems can be foreseen. Patients were included until 2012, limiting the ability to draw conclusions from more recent devices and in more EVAR experienced vascular surgeons. Another important issue to address is that there were no clear protocols as to when a patient might be discharged from further follow up. This could have modified the outcome that patients would automatically be considered as patients with (dis)continued follow up. However, this may affect both groups equally. No analysis was performed whether patients were treated in accordance with the IFU. Another major limitation of this study is the missing information on cause of death, and therefore the primary outcome regarding aneurysm related mortality should be interpreted with caution. Autopsy to confirm cause of death is not routinely performed in The Netherlands, and therefore it is likely that the proportion of aneurysm related deaths has been underestimated. Another limitation was that the reasons why patients were no longer under surveillance imaging were not noted, and therefore it can only be speculated what happened in the long interval between last imaging and time of death. This may have resulted in a distorted view as patients could have been referred back to their initial hospital, and therefore be misclassified as discontinued to imaging follow up.
The strengths of this study include the use of population based data with long term follow up and the accuracy of verifying deaths through the National Death Register. It also comprises all imaging studies made following EVAR, and thus provides a complete overview of national practice and adherence to the ESVS guidelines.
The results show that no EVAR related abnormalities were detected during follow up in the majority of patients (74%). In this multicentre study, discontinued imaging follow up was not associated with worse outcomes. This signals an important opportunity for increasing the efficiency of the post-EVAR follow up programme. Yearly imaging surveillance may not always be necessary if initial post-operative CTA shows no abnormalities. This suggestion should be validated in future prospective studies to determine in which patients imaging surveillance can be safely postponed. Incorporating risk stratification based on pre-operative variables into the development of surveillance protocols may reduce overtreatment. Since patients with no abnormalities at their initial post-EVAR CTA may be considered as low risk, imaging surveillance could be delayed until five years after the initial repair, as proposed by the ESVS guidelines. This would save on healthcare system resources without endangering patient safety, while at the same time increasing adherence among those who do need surveillance.
Conflicts of interest
None.
Funding
This study was funded by The Netherlands Organisation for Health Research and Development (ZonMw; grant number 843004119).
Acknowledgements
The authors thank N. Wolfhagen for the statistical assistance with Rstudio.
Appendix A. Supplementary data
The following is the Supplementary data to this article:
Endovascular versus open repair of abdominal aortic aneurysm in 15-years’ follow-up of the UK endovascular aneurysm repair trial 1 (EVAR trial 1): a randomised controlled trial.
Adequate seal and no endoleak on the first computed tomography angiography as criteria for no additional imaging up to 5 years after endovascular aneurysm repair.
Postoperative surveillance and long-term outcome after endovascular aortic aneurysm repair in the Netherlands: study protocol for the retrospective ODYSSEUS study.
We have read the paper by Geraedts et al. with interest.1 The (lack of) efficacy of image follow up after endovascular aneurysm repair (EVAR) remains an unresolved problem and this study from a large collaboration in The Netherlands (ODYSSEUS study) aimed to determine the result of non-compliance with existing protocols. All clinicians involved in the care of EVAR patients are aware of the burden of surveillance, and probably unsure of the efficacy of current strategies and consequences of non-adherence.
To submit a comment for a journal article, please use the space above and note the following:
We will review submitted comments as soon as possible, striving for within two business days.
This forum is intended for constructive dialogue. Comments that are commercial or promotional in nature, pertain to specific medical cases, are not relevant to the article for which they have been submitted, or are otherwise inappropriate will not be posted.
We require that commenters identify themselves with names and affiliations.
00905-9&id=gr1.jpg){kind=link}
00905-9&id=gr2.jpg){kind=link}