Since first reported in 1991 [1], endovascular aneurysm repair (EVAR) has emerged as an effective therapeutic modality for managing abdominal aortic aneurysms (AAAs). However, concerns persist regarding late complications following EVAR, most notably the enlargement of the aneurysmal sac and the potential secondary sac rupture [2]. Although sac expansion after EVAR is typically attributed to the presence of endoleaks, there are also occurrences of aneurysmal sac enlargement in the absence of a clearly defined endoleak. This phenomenon is referred to as “endotension” [3].

Several potential mechanisms contributing to endotension have been proposed, including pressure transmission through the stent-graft wall and transudation of fluid through the graft fabric. There is also a series of studies reporting that endoleak, previously undetected in prior images, was found during open conversion. However, the precise causative factors of endotension remain elusive [4-6].

The absence of a standardized treatment guideline for endotension poses a significant challenge in the management of affected patients. Various treatment options have been considered, taking into account factors such as sac size, growth rate, presence of symptoms, and potential mechanisms contributing to sac enlargement. However, the most effective treatment approach remains unclear. This study aimed to investigate and summarize the treatment of endotension in a single-center experience.

1) Patient selection

This study involved a retrospective review of single-center data from patients diagnosed with endotension following EVAR between January 2006 and December 2017. Fifteen patients were diagnosed with endotension among a total of 792 EVARs in the same study period. For each patient, the demographics, anatomic characteristics of the initial AAA, details of the initial EVAR procedure, presence of endoleak and endotension in follow-up, and their respective treatments were evaluated. The study protocol was approved by the Institutional Review Board of Asan Medical Center (IRB no. 2013-1027), waiving the requirement for written informed consent due to its retrospective nature.

2) Initial EVAR procedure

EVAR was performed by one vascular surgeon and one interventionist in this study population. The choice of stent-graft was determined based on an assessment of aneurysm morphology and physician preference. The anatomical factors attributing to the procedure of EVAR, including maximal diameter of aneurysm, proximal neck length and angle, and aneurysmal involvement of iliac arteries were collected in a retrospective review of computed tomography (CT) data. Some of the patients in our study population were referred to our center after an initial outside EVAR procedure due to the failure or complication of EVAR. In such cases, records from the referring hospital were also reviewed to compile data regarding the initial EVAR.

3) Clinical follow-up and diagnosis of endoleak and endotension

Routine follow-up of EVAR included CT angiography at 6 and 12 months after the procedure, with subsequent annual follow-ups. In patients with renal insufficiency, non-enhanced CT was utilized to assess the morphology of the aneurysm and, if necessary, Doppler ultrasound was employed for observing endoleak. The decision to use contrast in CT or angiography was made on an individual basis, taking into consideration the balance between benefits and risks.

For patients who exhibited endoleak, re-intervention was conducted based on the following indications: i) the presence of a type I or type III endoleak, and ii) a persistent type II endoleak with sac size increase of more than 5 mm compared with the baseline diameter. After the re-intervention, disappearance of endoleak was confirmed either by final angiography during the procedure or perioperative CT angiography. The determination of endoleak presence was crucial in the diagnosis of endotension. Multiple diagnostic modalities, including CT angiography with delayed phase images and Doppler ultrasonography, were employed to confirm the absence of endoleak.

4) Classification of the primary and secondary endotension groups

For evaluating the relation between prior endoleak and endotension, we divided the patients into two categories, primary and secondary, based on the presence of endoleak before the endoension diagnosis. The primary endotension (PE) group comprised patients diagnosed with endotension without any prior history of endoleak. In contrast, the secondary endotension (SE) group encompassed patients previously diagnosed with endoleak that resolved before the endotension diagnosis.

5) Outcome: treatment of endotension

Endovascular interventions for preemptive purposes were contemplated for patients without angiographic evidence of endoleaks. These interventions encompassed stent-graft relining, placement of an aortic cuff, or the extension of a limb graft to address suspected sources of pressure. The transmission of pressure through either the proximal or distal landing zone [7], including a possible undetected type I endoleak, was addressed by the strategic placement of an aortic cuff or the addition of an extra limb graft, respectively. In cases where no specific pressure source was suspected, stent-graft relining was contemplated to mitigate fluid exudation through the graft wall, a phenomenon commonly referred to as graft sweating. Direct sac puncture and aspiration, as a non-operative intervention, were also considered for pressure measurement and to assess the characteristics of the fluid within the sac.

The open surgical conversion was considered along with the failure of re-intervention for endotesion, the presence of symptoms, or the physician’s decision, taking into consideration the size of the aneurysmal sac. During the open conversion, the sac was punctured to measure the pressure following the preparation for aortic clamping. Subsequently, the sac was opened without aortic clamping to identify bleeding or other sources of pressure. The removal of the stent-graft with prosthetic graft interposition was mainly performed. In a subgroup of patients, marsupialization of the aneurysmal sac or proximal banding with subsequent relining of stent-graft was carried out based on the physician’s decision.

1) Clinical and anatomical characteristics

Baseline characteristics and an overview of general information about the aneurysms are presented in Table 1. All patients were male, and the mean age was 70.9 years. Nine patients had hypertension and five had chronic kidney disease at the time of EVAR. Referring to Supplementary Table 1, the Endurant device (Medtronic) was the most commonly employed stent-graft, followed by the Zenith device (Cook medical). The mean maximal aneurysm diameter was 64.3 mm.

Table 1 . Baseline characteristics of study population.

Characteristics	Findings
Male (sex)	15 (100)
Age (y)	70.9±4.3
Comorbidities
Hypertension	9 (60)
Diabetes mellitus	2 (13)
Coronary artery disease	1 (6)
Chronic kidney disease	5 (33)
Anatomic features of aneurysm
Maximal diameter (mm)	64.3±9.3
Involvement of iliac arteries	4 (27)
Proximal neck length (mm)	39.7±18.0
Proximal neck angle (°)	55.9±22.5

Results are presented as n (%) or mean±standard deviation..

2) Clinical information about endotension

Clinical information considering endotension is shown in Table 2. Among the study population, eight patients were classified into the PE group, while the remaining seven were in the SE group. The average duration from the last endovascular intervention for an aneurysm to the diagnosis of endotension was 25.7 months. The average duration from the diagnosis of endotension to the last follow-up was 30.5 months, during which the aneurysm size changed by approximately 20.3 mm due to endotension. Among 15 patients, four received direct open conversion and eight received preemptive endovascular intervention for the management of endotension. The preemptive intervention consisted of re-lining of the stent-graft in two patients, aortic cuff or limb graft extension in four patients, and aspiration of sac fluid with glue embolization inside the aneurysm sac in two patients. In three patients, active intervention for endotension was not conducted because of patients’ refusal and significant comorbidities; the aneurysm size continuously increased to nearly 100 mm over an average follow-up period of 45.6 months (Table 2).

Table 2 . Clinical information in patients with endotension.

Group	Patient	Duration (mo)			Diameter of aneurysmal sac (mm)		Preemptive intervention	Open conversion
		Before endotensiona	After endotensionb		Diagnosis of endontesionc	Finald
1°	1	36	15		74	97	Relining of stent-graft	Yes
	2	19	16		61	75		Yes
	3	26	43		69	114	Aortic cuff, extension of limb graft	Yes
	4	19	35		73	91	Relining of stent-graft	Yes
	5	17	36		55	74	Extension of limb graft	No
	6	31	39		50	73		Yes
	7	41	74		69	81	Extension of limb graft	No
	8	27	9		62	75	Extension of limb graft	No
2°	9	3	3		70	82	Aspiration	Yes
	10	1	28		77	94		Yes
	11	3	15		74	82		Yes
	12	60	39		93	110		No
	13	17	8		60	72	Aspiration	No
	14	48	59		49	104		No
	15	38	39		69	85		No

^aThe time from endovascular aneurysm repair or the last intervention for endoleak in the secondary endotension group until the diagnosis of endotension. ^bThe time from the diagnosis of endotension until the date of last follow up or surgery. ^cMaximal diameter of aneurysmal sac at the diagnosis of endotension. ^dMaximal diameter of aneurysmal sac in the last follow-up images..

3) History of endoleak in secondary endotension group

In particular, within the SE group (N=7), three patients had type Ia endoleak, two had type II endoleak, and two had type III endoleak (Table 3). To address these specific complications, endovascular interventions were individually tailored and performed on six patients, matching the type and characteristics of the identified endoleaks. Notably, in one patient with type Ia endoleak immediately after EVAR, open conversion with a proximal banding procedure was performed. All patients in the SE group showed resolution of endoleaks confirmed by conventional angiography or CT angiography after endovascular intervention or proximal banding. The time between EVAR, endoleak, and the diagnosis of endotension varied widely and lacked consistency. On average, it took 27.3 months from EVAR to the last endoleak intervention, and 24.3 months from the last endoleak procedure to the diagnosis of endotension. The diagnosis of endotension occurred at an average of 50.6 months from the initial EVAR.

Table 3 . Type of endoleak and its treatment in secondary endotension group.

Patient	Sex/age	Type of endoleak	Number of endovascular procedure (n)	Duration from EVAR (mo)
				EL+a	EL–b
9	M/69	Ia	1	3	88
10	M/78	Ia	1	1	1
11	M/67	III	1	3	17
12	M/69	III	2	60	50
13	M/75	II	1	17	5
14	M/74	Ia	0c	48	1
15	M/68	II	1	38	29

M, male; EVAR, endovascular aneurysm repair; EL, endoleak..

^aDuration from the last intervention for endoleak to the diagnosis of endotension. ^bDuration from EVAR to the last intervention for endoleak. ^cOne patient underwent open surgery (proximal banding) for type Ia endoleak, not endovascular intervention..

4) Operative findings and postoperative results

Table 4 presents clinical data for eight patients who underwent open conversion for the treatment of endotension during the follow-up period. In the majority of cases, the primary indication for surgery was sac enlargement. In one case, there was suspicion of a type III endoleak, indicated by the near separation of the main body and limb graft. Two patients required surgery due to an infected stent-graft during the surveillance of endotension, revealing a fistula with the aneurysm sac adjacent to the colon or appendix. These two patients were managed with explantation of the stent-graft, extensive debridement, complete isolation of any fistula, and omental wrapping. In the remaining five patients, serous to brownish fluid or old thrombi were observed inside the aneurysmal sac without evidence of endoleak. Partial explantation was the primary surgical treatment method for the removal of the stent-graft at our institution. Postoperative complications, including pneumonia, limb-graft occlusion, and sepsis were observed in three cases with no instances of 30-day mortality. Fig. 1 displays the Kaplan–Meier plot comparing survival based on surgery in patients with endotension, with no significant difference observed between the two groups.

Table 4 . Clinical information about open conversion in patients with endotension.

Patient	Group (°)	Indication of open conversion	OP name	OP finding	Complications
1	1	Sac growth	Proximal banding, distal fixation	Serous fluid	Graft occlusion
2	1	Sac growth	Explantation	Type III endoleak with the sac pressure of 60 mmHg
3	1	Sac growth	Explantation	Serous fluid
4	1	Infected stent-graft	Explantation	Fistula with sigmoid colon	Sepsis
6	1	Infected stent-graft	Explantation	Fistula with appendix
9	2	Sac growth	Explantation	Dark brownish fluid
10	2	Sac growth	Explantation	Dark brownish fluid	Pneumonia, ileus
11	2	Sac growth	Proximal banding, relining of stent-graft	Serous fluid

OP, operative procedure..

Figure 1. Kaplan–Meier plot for survival analysis based on open conversion.

In this retrospective review of endotension data, approximately 50% of patients who underwent preemptive endovascular intervention for endotension had shown no further progression of the size of the aneurysmal sac. In eight patients who underwent open conversion for endotesion, only one had bleeding inside the sac which is suspected of type III endoleak, while the others had serous or dark brownish fluid inside the sac. Approximately one-third of patients who received open conversion showed perioperative complications, and there was no operative mortality.

Late complications of EVAR, along with the need for re-intervention, have been prominent concerns in this field [2,8-11]. While endoleaks have received significant attention in previous studies, little was known about endotension until recent years, primarily due to its rarity. The incidence of endotension in earlier studies has been reported to range from 1.5% to 5% [12-15]. In our study group, the incidence of endotension was 3.28% (15 out of 792 patients), closely aligning with the findings of previous research.

The precise mechanisms underlying endotension remain elusive, although earlier studies have proposed several possible contributing factors, as previously mentioned. In line with previous experimental reports and case studies [4-6,16], recent review data also suggested potential causes of endotension with a similar mechanism [17]. According to this study, a plausible cause for endotension could be the pressure through the thrombus in the landing zone or at the orifices of the aorta and iliac artery branches, as well as the undetected endoleaks. However, highly evidenced data on the mechanism of endotension are still lacking.

We divided our study population into two groups based on the presence of a prior endoleak to investigate the relationship between endoleak and subsequent endotension. Among the patients in the SE group who underwent surgery, two patients received surgery relatively shortly after the treatment of endoleak with the possibility of undetected endoleak as a plausible cause of sac enlargement. However, it is noteworthy that even in these patients, surgical findings did not reveal the presence of endoleak. Moreover, bleeding was confirmed within the aneurysmal sac in only one patient, classified as a type III endoleak [18]. Referring to recent studies on endotension or late complications of EVAR, endoleak was confirmed during surgery for endotension in 75% and 30% of patients, respectively [15,19]. In contrast, this study observed a relatively low incidence of undetected endoleak in only one (12.5%) patient, with the remaining patients exhibiting endotension suspected to result from mechanisms other than bleeding.

The presence of serous or dark brownish fluid inside the aneurysmal sac has been indicative of transudate secretion or pressure transmission through the stent-graft [20]. In the current study, this type of fluid was observed in over 80% of patients who underwent open conversion accompanied by endotension, suggesting a similar underlying mechanism. In the early reports, conservative treatment was contemplated in these patients with meticulous selection [21].

In this study, preemptive procedures in three patients demonstrated efficacy in preventing subsequent sac size increases. Through a thorough review of serial images and medical records, these patients exhibited a high suspicion of pressure sources, such as stent-graft migration or the development of a saccular aneurysm in the distal common iliac artery. The absence of a size increase after corrective procedures suggest the possibility of a problem at the landing zone, including an undetected endoleak. Consequently, preemptive intervention emerges as a valuable consideration, particularly in select patients.

The enlargement of an aneurysmal sac can give rise to several complications. The accumulation of fluid within the aneurysmal sac has the potential to alter the aneurysm’s morphology, potentially leading to stent-graft migration and type I endoleak. Additionally, the continuous friction between the wall of the enlarged aneurysm and adjacent visceral organs could result in the formation of a fistula, representing another serious complication. Although there is no standardized indication for surgery in patients with endotension, certain conditions, such as the presence of symptoms, stent-graft migration in follow-up images, or an infected stent-graft, can warrant open conversion, not merely the enlarged sac itself [22,23]. In the context of open conversion procedures, explantation was performed in six out of eight patients, with one-third experiencing perioperative morbidity, while no postoperative mortality was reported. Notably, in a prior systematic review, a notable increase in late open conversions flowing EVAR was observed, and this procedure exhibited relatively low mortality [24]. As alternative surgical approaches, sacotomy with endograft preservation is a viable option [25], as indicated by recent studies, despite the absence of long-term follow-up data including the re-expansion rate of the aneurysmal sac [26].

This study has several limitations. Firstly, it is a single-center study, and the relatively small study population presents challenges for comprehensive statistical analysis. However, it is important to note that given the rarity of endotension, prior studies have been predominantly case reports or retrospective studies with small populations. Secondly, we encountered incomplete follow-up data. Specifically, in seven out of 15 cases, the cause of death remains unknown because these patients did not continue their follow-up at our center. Some patients opted for outpatient care at hospitals in their province.

In the case of endotension, physicians should consider open surgery when patients exhibit symptoms or when stent-graft migration is suspected in serial follow-up images. In the absence of these indications, close observation and preemptive intervention for potential pressure sources can be considered for selected patients. The explantation with graft interposition can be a principal option for surgical intervention for endotension.

Supplementary Table can be found via https://doi.org/10.5758/vsi.230108

None.

Jun Gyo Gwon and Youngjin Han have been the editorial board members of the VSI since 2019. They were not involved in the review process. Otherwise, no potential conflict of interest relevant to this article was reported.

Conception and design: TWK, YPC, JGG, YH. Analysis and interpretation: JYK. Data collection: JYK, SAL. Writing the article: JYK. Critical revision of the article: TWK. Final approval of the article: all authors. Statistical analysis: none. Obtained funding: none. Overall responsibility: TWK.