Abstract
Our objective was to determine the relative merits of intervention or observation of type II endoleaks (T2Ls). A retrospective analysis was performed on 386 infra-renal endovascular aneurysm repair (IR-EVAR) patients from 2006 to 2015. Annual surveillance imaging of patients undergoing EVAR at our centre were analysed, and all endoleaks were subjected to a multidisciplinary team meeting for consideration and treatment. In the 10-year time frame, 386 patients (79.5±8.7 years) underwent an IR-EVAR. Eighty-one patients (21.0%) developed a T2L and intervention was undertaken in 28 (34.6%): 17 (60.7%) were treated via a transarterial approach (TA) and 11 (39.3%) using the translumbar approach (TL). Fifty-three patients (65.4%) with T2Ls were managed conservatively. Patients who received T2L treatment had a greater proportion of recurrent T2Ls than patients who were conservatively managed (p=0.032). T2Ls associated with aneurysmal growth were more resistant to treatment than those where there was no change or a decrease in aneurysm size during follow-up (0.033). There was no significant difference in the TA and TL approach with respect to endoleak repair success (p=0.525). Treatment of a T2L did not confer a survival advantage compared to conservative management (p=0.449) nor did the choice of either the TA or TL approach (p=0.148). Our study suggests the development of a T2L associated with aneurysm growth may represent an aggressive phenotype that is resistant to treatment. However, this did not lead to an increased risk of mortality over follow-up. Neither a transarterial nor a translumbar approach to treating a T2L conferred superiority.
Introduction
Abdominal aortic aneurysm (AAA) is a common disease with an estimated prevalence of 244.54 per 100,000 in western Europeans.1,2 As the aortic diameter approaches 50–55 mm, surgical intervention is usually undertaken as the risk of rupture is believed to outweigh perioperative risks for most patients. 3 Conventional open surgical repair was the treatment for AAA for more than half a century 4 but it has been shown to have a higher perioperative mortality than endovascular aneurysm repair (EVAR). 5 Thus, in the modern era, EVAR has replaced conventional surgery as the standard of care for patients with suitable anatomy.
The development of endovascular techniques has also introduced new postoperative complications, including endoleaks. Type II endoleaks (T2Ls) occur when retrograde blood flow from aortic collateral arteries pressurize the aneurysm sac. The culprit vessel is commonly a lumbar artery or the inferior mesenteric artery (IMA) after infra-renal aneurysm repair. The natural history of T2Ls is yet to be fully understood: some studies suggest T2Ls are a benign condition that resolves spontaneously without intervention 6 while others associate T2Ls with adverse outcomes such as aneurysm expansion and rupture.7,8 Although most agree that endoleaks in the setting of sac expansion are an indication for intervention, an optimal management strategy has not yet been established.9,10 When performed with endovascular techniques, T2L repair involves embolization of feeding vessels via a transarterial (TA) approach or with direct sac puncture using the translumbar (TL) approach and filling the sac with coils or glue. 11
In this study, we describe one centre’s experience with T2Ls post infra-renal EVAR (IR-EVAR) between 2006 and 2015. The clinical characteristics between patients who developed, and those who did not develop, a T2L is described, as well as outcomes comparing different types of treatment. The two approaches to T2L repair are also compared and survival analysis between these groups is analysed.
Method
Patients undergoing IR-EVAR were identified by a review of the national registry data from our centre between 2006 and 2015. Baseline preoperative demographic and clinical characteristics, including sex, age, smoking status, ASA score, admission status and preoperative maximum aneurysm diameter, were recorded. Follow-up data, including maximum aneurysm diameter, aneurysm sac size change, reintervention details, incidence of ruptures and additional endoleaks, were also collected from medical records. This study did not fall under the remit of the NHS Research Ethics Committee according to guidance from the Health Research Authority.
All patients undergoing EVAR at our centre are invited to annual surveillance imaging follow-up. During the standard course of follow-up, any suspected endoleak was reported and classified according to the reporting standards delineated in Chaikof et al. 12 An endoleak was defined by the persistence of blood flow outside the lumen of the endoluminal graft but within the aneurysm sac, as determined by arterial duplex and contrast-enhanced computed tomography (CT) imaging. During this time period, there was no systematic approach to endoleak management, but all endoleaks were subjected to a multidisciplinary team meeting for consideration and treated according to the wisdom of the group upon individual review. The clinical management of T2Ls consisted of either conservative watchful waiting or T2L repair; the direction of care was chosen after considering growth, patient comorbidities and likelihood of success.
T2L repair was classified as early if undertaken within 30 days of detecting a T2L and late if undertaken after more than 30 days. T2L repair consisted of either TA or TL embolization. Initially, CT angiography is obtained to determine if there is an endovascular pathway to the aneurysm sac. As the majority of T2Ls occur from the inferior mesenteric artery (IMA) or lumbar arteries, the superior mesenteric artery (SMA) to IMA and internal iliac to lumbar collateral pathways are examined. If an appropriate route is identified, the transarterial approach is attempted. The SMA or internal iliac arteries are selected with a 5 Fr catheter. A microcatheter and microwire are then used to navigate the collateral pathway to the feeding vessel and coils are deployed within the vessel.
If the transarterial approach is not feasible or fails, a percutaneous translumbar approach is utilized. The patient is placed prone and, under fluoroscopic guidance using bone and metallic landmarks, an 18-gauge to 20-gauge needle is used for access into the sac. Once blood return is achieved, positioning within the sac is confirmed with contrast. Coils or an embolizing agent is then injected within the sac.
Patients were followed up with regular CT imaging and a T2L was deemed successfully treated at the point where there was no endoleak detectable and/or the AAA sac size reduced. The reappearance of an endoleak after spontaneous resolution or after a repair that was considered successful is defined as a recurrent endoleak. 12
Changes in aneurysm dimensions are reported according to the standards set out by Chaikof et al. 12 The preoperative maximum aneurysm diameter is recorded as the diameter measurement before the EVAR. Follow-up maximum aneurysm diameter is recorded as the most recent diameter measurement of the aneurysm sac. Aneurysm growth was any increase in diameter by 5 mm or more, and a decrease in aneurysm diameter by 5 mm or more was also noted. An absence of significant change in AAA diameter was noted as no change in aneurysm diameter. 12
Statistical methods
Chi-squared and Fisher exact tests were used to compare demographic and clinical characteristics between the subgroups. Age, preoperative AAA size and follow-up AAA size were not normally distributed and thus non-parametric Wilcoxon signed rank and Mann–Whitney U tests were used to determine whether any statistically significant difference existed between the comparison groups. Survival analysis was performed by the Kaplan–Meier method to estimate survival function stratified by T2L status, T2L repair and T2L repair type. The log-rank test was used to elucidate any statistical difference in the survival functions among the subgroups. All statistical analyses were performed using SPSS for Mac (version 24; IBM Corporation, Somers, NY, USA) at the significance threshold of p<0.05.
Results
During the 10-year inclusion period (2006–2015), 386 patients underwent elective and emergency IR-EVAR. The medical records and imaging studies for these patients were retrospectively reviewed. The median follow-up time was 3.11 years (interquartile range (IQR), 1.52–5.53 years). A T2L developed in 81 (21.0%) patients, of which 28 (34.6%) were surgically repaired: 17 (60.7%) were repaired using the TA approach and 11 (39.3%) using the TL approach. Fifty-three (65.4%) of the 81 T2Ls were conservatively managed. Over the course of follow-up, 27 (7.0%) of the 386 patients developed a type I endoleak, three (0.8%) developed a type III endoleak and three (0.8%) developed a combination of both type I and III endoleak.
Baseline demographics and clinical characteristics (Table 1)
Baseline preoperative demographic and clinical characteristics of 386 infra-renal endovascular aneurysm repair (IR-EVAR) results.
Patients either died before follow-up computed tomography or duplex imaging, or they were transferred to another hospital.
Comparison between the T2L and no T2L groups.
Comparison by Wilcoxon–Mann–Whitney test.
Baseline demographic and clinical characteristics of patients who developed a T2L (n=81) and those who did not (n=276) are listed in Table 1. The mean age of patients who underwent an IR-EVAR was 79.5±8.7 and 83.9% were male. Comparing those who went on to develop a T2L with those who did not, there was no significant difference in the preoperative maximum aneurysm diameter (p=0.121) but there was a significant difference in the end of follow-up aneurysm dimensions. Patients who developed a T2L had a larger end of follow-up aneurysm diameter (63.0±19.0 mm) compared with those who did not develop a T2L (53.4±14.2 mm) (p<0.001). Furthermore, the aneurysm sac decreased in 72.1% of patients who did not develop a T2L compared with 43.2% of patients who did develop a T2L (p<0.001).
Additional endoleaks (type I or type III) occurred in 33 (8.5%) of the 386 patients. However, there was no significant difference between the incidence of an additional endoleak in patients who developed a T2L and in patients who did not (p=0.155). All other baseline demographics and clinical characteristics between the two subgroups were similar (Table 1). These included sex (p=0.094), age (p=0.421), smoking status (p=0.128), ASA score (p=0.714) and EVAR admission status (p=0.951).
Type II endoleak repair compared to conservative management (Table 2)
Clinical characteristics of 81 patients who developed a type II endoleak post infra-renal endovascular aneurysm repair (IR-EVAR) stratified by treatment.
Comparison between treated and not treated T2L.
Comparison result by Wilcoxon–Mann–Whitney test.
Table 2 compares the clinical outcomes of T2L patients (n=81) who received T2L treatment (n=28) to those managed conservatively (no reintervention) (n=53). There was a difference between T2Ls that were treated as opposed to T2Ls not treated. Patients who received T2L treatment had a greater proportion of recurrent T2Ls (reappearance of an endoleak after spontaneous resolution or after a repair that was considered successful) 12 than patients who were conservatively managed (p=0.032). Moreover, 64.3% of treated T2Ls had an increase in aneurysm sac diameter during follow-up compared with 41.5% of non-treated T2Ls (p=0.132). It was found that 22.2% of T2Ls associated with an increase in aneurysm sac diameter had a recurrence of a T2L after treatment. Whereas, none of the T2Ls in which the aneurysm remained static or decreased in size had a T2L recurrence post repair (p=0.033). All other clinical and patient characteristics including age (p=0.210), sex (p=0.282) and EVAR admission status (p=0.409) were statistically similar between those treated and those conservatively managed (Table 2).
Transarterial approach compared to the translumbar approach (Table 3)
Clinical characteristics of 28 patients treated for a type II endoleak post infra-renal endovascular aneurysm repair (IR-EVAR).
Comparison between TA and TL approach.
Comparison result by Wilcoxon–Mann–Whitney test.
Of the 28 T2Ls treated, 17 (60.7%) were treated using the TA approach and 11 (39.3%) using the TL approach (Table 3). There was no significant difference between the two treatment methods with respect to the proportion of successful T2L treatments (a decrease in aneurysm sac size and/or a resolved endoleak at the end of follow-up): clinical success was achieved in 11.8% of patients who underwent the TA approach and in 27.3% of patients who had the TL approach (p=0.525). The preoperative AAA size in the TA approach (60.6±18.9) was statistically similar to the preoperative AAA size in the TL approach (65.8±9.6) (p=0.259). The same is true for the end of follow-up AAA size (p=0.603). In terms of aneurysm sac size change, there was no significant difference between the TA and the TL approach (p=0.471). There was no difference in any other measured metric (Table 3).
AAA ruptures
There were no AAA ruptures post EVAR reported in this study.
Time to event analysis
Time to event analysis demonstrated patients with a T2L were more likely to survive than those who did not develop a T2L (p=0.029); see Figure 1 and Table 4. However, there was no survival difference between whether a T2L was treated or not (p=0.449) (see Figure 2 and Table 5), nor if the T2L was treated using the TA approach or the TL approach (p=0.148) (see Figure 3 and Table 6).
Effect of type II endoleak (T2L) on all-cause mortality. Kaplan–Meier curves for infra-renal endovascular aneurysm repair (IR-EVAR) patients (n=357) stratified by T2L status. Survival did differ between those with and without a T2L (p=0.029, log-rank test).
Associated Kaplan–Meier curve (Figure 1) case processing summary (n=357).
Effect of type II endoleak (T2L) repair on all-cause mortality. Kaplan–Meier curves for T2L patients (n=81) stratified by whether they received T2L treatment or were conservatively managed. Survival did not differ between those with and without a T2L (p=0.449, log-rank test).
Associated Kaplan–Meier curve (Figure 2) case processing summary (n=81).
Comparing the transarterial (TA) and the translumbar (TL) approach in terms of all-cause mortality (n=28). Survival did not differ between those with and without a T2L (p=0.148, log-rank test).
Associated Kaplan–Meier curve (Figure 3) case processing summary (n=28).
Discussion
This study describes a decade of experience with the management of T2Ls post IR-EVAR in a vascular centre. In our experience, 21.0% of patients develop a T2L during follow-up, which is consistent with past research which has shown T2L incidence to range from 7.8% to 44.0%.13–15 The length of follow-up in this study is 3.11 years (3.11 years [IQR 1.52–5.53 years]), and is longer than other studies in this field.16–18 Patients with T2L were less likely to have a decrease in aneurysm size during postoperative follow-up (p<0.001). Patients with T2L seem to have a survival advantage over patients who did not develop a T2L on time to event analysis (p=0.029) (Figure 1), but this is likely to be a type I error due to the low event rate in both populations. There is no difference in all-cause mortality in T2L patients compared with patients without T2L.
Our study suggests that T2Ls with findings that prompt clinical intervention at our institution – such as sac growth – are at higher risk of having recurrent endoleaks and continuing to grow. It is unclear from this analysis if this is because the treatments provided are ineffective, or if aneurysms that develop these aggressive T2Ls are by their nature more likely to have a recurrent blood supply. Our study found 22.2% of T2Ls with sac growth had a recurrence of an endoleak post repair. This is in comparison to no recurrence in aneurysms that remained static or decreased in size. Endoleaks persisted in 67.9% of patients who underwent at least one intervention for T2L, compared with 41.5% in patients who were treated conservatively with no intervention. The published literature has similar examples of the apparent ineffectiveness of T2L repair. Aziz et al. 19 observed recurrent T2Ls in 72.0% of patients who underwent treatment. Jouhannet et al. also reported a high failure rate for T2L treatment (60.0%) and called into question the high costs of such reinterventions. 20 Gelfand et al. found 58.0% of T2Ls which were left without treatment had resolved completely within 1 year post EVAR. 21 Our resolution rate for untreated T2Ls was 47.2% and 17.9% for T2Ls that were treated. These results suggest T2L patients who did not have treatment were more likely to have a resolved T2L. However, a major criticism of the conservative management of T2Ls is the notion that untreated T2Ls are more prone to rupture. 8 This study found no evidence of a T2L being associated with an AAA rupture and similar findings have been reported by Silverberg et al. 16 and Gelfand et al. 21 Moreover, Kaplan–Meier survival analysis found no difference in survival probability between whether a T2L was treated or not (p=0.449) (Figure 2).
Given that there is an aggressive phenotype that appears to be emerging about the characteristics of patients who develop endoleaks leading to sac growth, it is important to explore other physiologically plausible explanations. The behaviour of these endoleaks could be explained by vascular neogenesis. Certainly the arteriovenous malformation literature is recently exploring theories on the role of vascular endothelial growth factor (VEGF) blockade as a method to limit vascular neogenesis.22,23 If it were possible to borrow from these ideas, there may be a pharmacologic approach to type II endoleak that could be explored.
In terms of comparing the TA approach to the TL approach, we found no meaningful difference between them. Although a greater percentage of patients who received the TL approach achieved clinical success (27.3% compared with 11.8%), this was not significant (p=0.525) (Table 3). Both techniques offered similar AAA sac size changes and endoleak status at the end of follow-up. Our study found both approaches equally successful and this was in accordance with Stavropoulos et al. 2 who also found the TA and TL approach similar in terms of effectiveness. Survival analysis further demonstrated no significant difference between patients receiving either approach (p=0.148) (Figure 3).
Limitations
This is a retrospective study of one centre’s experience that spans numerous practices and generations of endovascular devices, so there are weaknesses that may limit the generalizability. Follow-up at a vascular centre where many patients return to the hospital of referral for their follow-up suggests that there may be events that have gone unrecorded in our records. Additionally, the process by which endoleaks were adjudicated and treatment was considered was highly dependent on the team responsible for the patients, and thus was not standardized. Overall, the event rate is low, which makes statistical tests of association weak, and, we agree, should be used for hypothesis generation only.
Conclusion
In conclusion, the development of a T2L associated with aneurysm growth may represent an aggressive phenotype that is resistant to treatment. However, this did not lead to an increased risk of mortality over the follow-up of this study. Neither transarterial nor translumbar approach emerged as the treatment modality of choice for repairing T2L.
Footnotes
Declaration of conflicting interests
The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Tara M Mastracci has consultation and speaking agreements with Maquet Getinge Group and Cook Medical Inc.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.