Catheter-based treatment of the dissected ascending aorta: a systematic review

Summary

OBJECTIVES

Type A aortic dissection requires immediate surgical repair. Despite improvements in surgery and anaesthesia, there is still a considerable risk when high-risk patients are concerned. Less invasive endovascular treatments are under evaluation. We investigated the current status of catheter-based treatment for type A aortic dissection with the entry tear located in the ascending aorta.

METHODS

A PubMed search was supplemented by searching through bibliographies and key articles. Demographics, risk score, stent graft detail, access route, mortality, cause of death, complications, reinterventions and follow-up data were extracted and analysed.

RESULTS

Thirty-one articles (7 retrospective reports; 24 case reports/series) were included in the study. In total, 104 patients (mean age 71 ± 14 years) received endovascular treatment for acute (63) or chronic (41) type A dissection. A history of a major cardiac or aortic operation was present in 29 patients. The mean EuroSCORE II was 30 ± 20 in 4 reports. A total of 114 stent grafts were implanted: ‘off-the-shelf’, 65/114; custom made, 12/114; and modified, 7/114. Hospital complications included intraprocedural conversion to open surgery (2/104), stroke (2/104), coronary stenting (2/104), early endoleak (9/104) and repeat aortic endovascular treatment for endoleak (5/104). Hospital mortality was 10% (intraoperative death 2/104). Mean duration of follow-up time was 21 ± 21 months (range 1–81 months); follow-up data were available for 86 patients: 10 patients died of non-aortic-related causes; reintervention for aortic disease (endovascular repair or open surgery) was performed in 8 patients.

CONCLUSIONS

Catheter-based ascending aorta repair for type A aortic dissection with the entry tear in the ascending aorta can be considered in carefully selected high-risk patients. Further analysis and specifically designed devices are required.

Endovascular aorta repair, Catheter-based techniques, Ascending aorta dissection

INTRODUCTION

The recommended treatment for type A aortic dissection (TAAD) is the surgical repair of the ascending aorta, with or without arch replacement, and the frozen elephant trunk technique. However, despite improvements in surgical and anaesthesiological protocols and techniques, operative mortality still ranges between 10% and 25%, and the overall mortality remains around 20–30%, according to the International Registry of Acute Aortic Dissections report and published series [1–4]. Moreover, the standard technique shows worse results in older patients with severe comorbidities. Approximately 20% of all patients with acute and chronic TAAD are deemed not suitable for standard repair by the surgeons because of their advanced age and comorbidities or they are not operated on because the patient refuses to give consent because of fear of death or of complications such as neurological events and permanent dialysis [3, 4]. These patients are managed medically, which is associated with poor outcome.

As a minimally invasive alternative treatment for patients with high-risk profiles, catheter-based endovascular repair of the ascending aorta is carefully and slowly evolving. The use of this technique in type A dissection or in other serious diseases of the ascending aorta has been reported [5, 6]. However, the suitability of the ascending aorta for an endovascular repair is still limited by its complex anatomy and functional properties and by the technical challenge related to the absence of specifically designed endografts. According to the guidelines of the European Society of Cardiology, endovascular therapy for TAAD can be attempted in selected high-risk cases but has not yet been validated [7]. Our goal was to analyse the current status of catheter-based techniques for the treatment of TAAD with the primary entry tear located in the ascending aorta.

METHODS

This systematic review was performed and reported in line with the preferred reporting items for systematic reviews and meta-analyses statement [8]. We searched MEDLINE (https://www.ncbi.nlm.nih.gov/pubmed/) through the end of December 2019 using medical subject headings and text words supplemented by scanning the bibliographies of recovered articles. We combined ‘endovascular treatment’ and ‘ascending aorta’ using the Boolean operator ‘AND’. We used similar search strategies using the terms ‘endovascular repair’, ‘endovascular stent grafting’, ‘endograft repair’, ‘endovascular stent grafts’ and ‘proximal ascending aorta’, ‘ascending aortic dissection’, ‘type A aortic dissection’. Language was limited to articles written in English. Two co-authors (C.W. and L.K.v.S.) reviewed and selected relevant articles for inclusion. Differences were resolved in consensus discussions.

Inclusion criteria

We used ‘endovascular treatment of the dissected ascending aorta’ as the initial retrieval. Only cases receiving endovascular stent grafts for the treatment of a confirmed dissected ascending aorta (acute or chronic) with the primary entry tear located in the ascending aorta were considered. Only papers reporting overall morbidity and mortality data, definitive treatment modalities and configuration of catheter-based techniques were included.

Exclusion criteria

All research or case studies reporting on the treatment of non-dissected ascending aortic diseases (ascending aorta aneurysm, pseudoaneurysm, penetrating ulcer and thrombosis) were excluded. Reports on hybrid operations, the primary entry tear located in the aortic arch or in the descending aorta or retrograde type A dissections generated from type B treated with thoracic endovascular aortic repair in the descending aorta were also excluded. Correspondence, expert opinions and reviews were also excluded.

Data extraction

We collected data on study design (retrospective study, case series and case report), age, gender, case numbers, type of dissection (acute or chronic), preoperative risk score (EuroSCORE, Society of Thoracic Surgeons score, Japan score and American Society of Anesthesiologists score), details of the stent graft, access route, mortality, cause of death, complications, early and late aortic reinterventions, follow-up length and long-term outcome.

Statistical analyses

The collected data were organized and analysed with an Apple Numbers programme (version 6.6.2). Descriptive statistics was used to describe demographic data and stent graft details. Continuous variables were described with the mean ± standard deviation whereas dichotomous variables were expressed as the absolute number with percentage.

RESULTS

The literature search yielded 4590 publications in PubMed. We screened them by title/abstract and full text. Finally, 31 publications focusing on catheter-based treatment of the dissected ascending aorta with the primary entry tear in the ascending aorta were identified from the English-language literature and included in the analysis, spanning a period from 2000 to 2019 (Fig. 1).

Figure 1:

Flow diagram to illustrate identification, selection and exclusion of published articles used for this review.

Demographics, indications and outcome

All data are shown in Table 1. There were 7 retrospective reports and 24 case reports or small case series [6, 9–38]. The sample size ranged widely from 1 to 16 patients. A total of 104 patients were included in the reports. The mean age was 71 ± 14 years (range 41–95 years); 52 patients were men and 36 were women. In 1 study that included 16 cases, the gender was not disclosed. In the majority of the reports (90 cases), the indication for catheter-based TAAD treatment was the patient’s high-risk profile for standard open surgery. In 1 case, the percutaneous intervention was performed as a bailout procedure to stabilize the patient in the catheterization laboratory before proceeding with a standard operation [22]. TAAD were considered acute in 63/104 cases (61%) and chronic in 41/104 cases (39%) following the original article definition or, when stated, the temporal International Registry of Aortic Dissection classification (acute: 2–7 days; chronic: >8 days). Twenty-nine patients (28%) had a history of a major cardiac or aortic operation, and 5/104 underwent ascending aorta or root replacement. In particular, 1 patient had a residual dissection with a pseudoaneurysm after the aortic graft; 2 had an on-going dissection from the distal anastomosis after root replacement and ascending aorta replacement for TAAD; 1 had a distal dissection with a pseudoaneurysm after a Bentall procedure; and 1 had a dissection from the distal anastomosis of a Bentall procedure following iatrogenic dissection during coronary surgery [9, 10, 38].

Table 1:

Case series summary of endovascular treatment of the dissected ascending aorta

Author	Article types	Year of publication (data range)	Number of patients (/total)	Age (years)	Gender	Disease type following temporal IRAD classification (if applicable)	Previous cardiac or aortic operation	Indications	Risk score	Access route	Mean follow-up (months)	Hospital mortality	Cause of death	Late death
Li et al. [6]	Retrospective analysis	2016 (2009–2011)	15	65 ± 12 (range: 45–78)	M 12 F 3	aTAAD: 1 cTAAD: 14	None	Patients with multiple high-risk factors (multidisciplinary consultation board)	ASA score: IV or greater	TF	72 (range: 61–81)	0		0
Tsilimparis et al. [9]	Retrospective analysis	2019 (2010–2017)	16 (/24)	NA	NA	aTAAD: 8 cTAAD: 8	CABG: 1 AVR: 1 TAVI: 1 AAR: 1	Patients rejected for open surgery due to comorbidities, surgical history and/or age	ASA score: III–IV	Unspecified: TF, TA, TAX	11 (range: 0–35)	3 (19) (all aTAAD)	Type Ia endoleak: 1; sepsis: 1; unknown cause: 1	0
Grieshaber et al. [10]	Retrospective analysis	2018 (2014–2015)	6	66 ± 9 (range: 49–76)	M 5 F 1	aTAAD: 6	AAR: 3 AVR: 1	High estimated perioperative risk	EuroSCORE II: 34 ± 10 (range: 22–51)	TF + TA	23	0		0
Nienaber et al. [11]	Retrospective analysis	2017 (2009–2016)	12	81 ± 7 (range: 69–90)	M 9 F 3	aTAAD: 6 cTAAD: 6 (acute: <14 days from symptoms)	TAVI: 1	Unacceptably high surgical risk related to comorbidities	EuroSCORE II: 9 ± 4 (range: 3.9–19.3); ASA score: IV or greater	TF	21 ± 12 (range 0–36)	1 (8) (cTAAD)	Tamponade from guidewire-induced left ventricular perforation (intraoperative death)	4: non-aortic-related causes (15, 19, 23, 36 months later)
Roselli et al. [12]	Retrospective analysis	2015 (2006–2014)	13 (/22)	72 ± 12 (range: 51–91)	M 6 F 7	aTAAD: 11 cTAAD: 2	Unspecified: 7	Patients were considered at too high risk for conventional open repair	NA	TA: 6 TF: 6 TAX: 1	NA	3 (23) (aTAAD)	Multiple organ failure : 1; ventricular rupture and tamponade (intraoperative death): 1; late rupture and tamponade: 1	3: pneumonia, lung cancer and unknown cause
Bernardes et al. [13]	Retrospective analysis	2014 (2007–2012)	3 (/7)	54 ± 5 (range: 48–58)	F 3	aTAAD: 2 cTAAD: 1	CABG: 1	Patients were considered at high risk for open surgery by a multidisciplinary team	ASA score: III	TF	25 ± 21 (range: 3.5–46.4)	0		0
Ye et al. [14]	Retrospective analysis	2011 (2001–2009)	10 (/45)	51 (range: 41–71)	M 9 F 1	aTAAD: 6 cTAAD: 4 (acute: <14 days from symptoms)	NA	After discussion of the surgical options, the patient or relatives made the decision as to the type of surgery	NA	TC: 2 TF: 8	35 ± 5	1 (10)	Gastrointestinal bleeding	1: pneumonia (10 months later)
Wamala et al. [15]	Case report	2019	1	91	M	aTAAD	EVAR for AAA	Age and prohibitive risk for open surgery	NA	TA	12	0		0
Khoynezhad et al. [16]	Case report	2019	1	88	F	aTAAD (intramural haematoma)	None	Not an open surgical candidate due to age and comorbidities	NA	TF	NA	0		NA
Murakami et al. [17]	Case report	2017 (2013–2015)	1 (/6)	66	F	cTAAD	None	Frailty and disability after previous stroke	Japan score: 9.4	TA	NA	0		NA
Yang et al. [18]	Case report	2016	1	73	F	aTAAD	None	The patient refused surgery because of comorbidities	NA	TF	6	0		0
Khoynezhad et al. [19]	Case series	2016	2 (/6)	88, 84	M 1 F 1	aTAAD: 1 cTAAD: 1	Unspecified: 1	High-risk surgical candidates with ASA score of IV	ASA score: IV	TF	4	0		1: gastrointestinal bleeding (4 months later)
Vallabhajosyula et al. [20]	Case series	2015 (2007–2013)	2 (/6)	90, 78	F 2	aTAAD	None	Various factors rendered open surgical approach of high or prohibitive risk	NA	TA	9 ± 4	0		1: pneumonia (6 months later)
Wilbring et al. [21]	Case report	2015	1	82	M	aTAAD	None	The patient was denied surgery because of relevant comorbidities	NA	TF	6	0		0
Tauchi et al. [22]	Case report	2014	1	66	F	aTAAD	None	Decision made when haemodynamics deteriorated in catheterization laboratory due to coronary obstruction (bailout procedure)	NA	TC	NA	0		NA
Luo et al. [23]	Case report	2014	1	56	F	aTAAD	Heart transplant	High-risk redo case	NA	TA	6	0		0
Berfield et al. [24]	Case report	2014	1	95	F	aTAAD (3 days after TAVI)	TAVI	Prohibitive risk for surgery	STS score: 10%	TF	2	0		0
Ronchey et al. [25]	Case series	2013 (2009–2012)	4	70 ± 9 (range: 58–78)	M 2 F 2	aTAAD: 3 cTAAD: 1	AAR (aTAAD): 2 AVR: 1	Patients considered at high risk for open ascending aorta repair	EuroSCORE II: 29 ± 13 (range: 9.4–37) ASA score: III–IV	TF	15 (range: 4–39)	0		0
Shabaneh et al. [26]	Case report	2013	1	70	M	cTAAD	CABG; surgery for AAA; EVAR for AAA	Patient at high surgical risk (redo)	NA	TF	1	0		0
Pontes et al. [27]	Case report	2013	1	84	F	aTAAD	None	Conventional treatment unfeasible due to comorbidities	EuroSCORE II: 83	TF	NA	0		NA
Pinaud et al. [28]	Case report	2013	1	83	F	aTAAD	TEVAR for TAAA	Patient considered inoperable by multidisciplinary team	NA	TA	NA	0		NA
McCallum et al. [29]	Case report	2013	1	77	M	cTAAD	Heart transplant	High-risk case with 4 resternotomies and severe comorbidities	NA	TF	25	0		0
Metcalfe et al. [30]	Case report	2012	1	68	F	aTAAD	None	Patient considered unfit for surgery because of comorbidities	NA	TF	NA	0		NA
Ruchat et al. [31]	Case report	2009	1	64	M	aTAAD (iatrogenic during MVR)	Pericardiectomy; MVR	Increased aortic diameter after 3 months from MVR. Patient at high risk for redo surgery	NA	TC	12	0		0
Palma et al. [32]	Case report	2008	1	63	M	cTAAD	None	Patient at high surgical risk for conventional approach	NA	TF	0	1 (100)	Severe aortic regurgitation (probable SG migration)	0
Senay et al. [33]	Case report	2007	1	66	M	aTAAD	None	Conventional repair considered to be very high risk	NA	TF	NA	0		NA
Zimpfer et al. [34]	Case report	2006	1	84	M	aTAAD	None	Patient considered too high risk for surgery because of comorbidities	NA	TF	1	0		0
Zhang et al. [35]	Case report	2004	1	46	F	aTAAD	None	Patient refused surgery	NA	TF	22	0		0
Ihnken et al. [36]	Case report	2004	1	89	F	aTAAD	None	Elderly patient in overall frail status	NA	TF	NA	0		NA
Wang et al. [37]	Case report	2003	1	46	F	aTAAD (Marfan)	None	Not described	NA	TF	12	0		0
Dorros et al. [38]	Case report	2000	1	56	F	aTAAD	CABG + AAR (iatrogenic aTAAD)	High surgical risk	NA	TS + TF	1 (3 weeks)	1 (100)	Cardiac arrest (3 weeks later, after discharge)	0
Total	Retrospective analyses, 7 Case reports and case series, 24	2000–2019	104	71 ± 14 (range: 41–95; data available for 88 cases)	M 52 F 36 (NA: 16)	aTAAD: 63 cTAAD: 41	29 (28%) [5 patients (5%) had previous AAR]	High-risk cases: 90 Decision left to the patient: 10 Patient refused surgery: 2 Bailout procedure before standard surgery: 1 Unknown: 1	EuroSCORE II 30 ± 20 (3.9–83) (4 papers, 23 cases) ASA score: III–IV (5 papers, 52 cases) STS score: 10% (1 paper, 1 case) Japan score: 9.4 (1 paper, 1 case)	TF: 64/104 (73%) TA: 12/104 (14%) TC: 4/104 (4%) TAX: 1/104 (1%) TA + TF: 6/104 (6%) TS + TF: 1/104 (1%) Unspecified TA, TF, TAX: 16 cases	21 ± 21 (range 0–81) (data available for 84 cases)	10 (10%) (7 were aTAAD; intraoperative death: 2)		10/86 patients with follow-up (not aortic-related)

Values are expressed as n (%) or mean ± SD.

AAA: abdominal aortic aneurysm; AAR: ascending aorta replacement/repair; ASA: American Society of Anesthesiologists; aTAAD: acute type A aortic dissection; AVR: aortic valve replacement; CABG: coronary artery bypass grafting; cTAAD: chronic type A aortic dissection; EVAR: endovascular aortic repair; IRAD: International Registry of Aortic Dissection; MVR: mitral valve replacement; NA: not available; STS: Society of Thoracic Surgeons; SD: standard deviation; TA: transapical; TAAA: thoraco-abdominal aortic aneurysm; TAVI: transcatheter aortic valve implantation; Tax: transaxillary; TC: transcarotid; TEVAR: thoracic endovascular aortic repair; TF: transfemoral; TS: transseptal.

The mean EuroSCORE II was 30 ± 20 and was reported in 4 studies (range 3.9–83); the American Society of Anesthesiologists score was III–IV in 6 studies; 1 paper reported a Society of Thoracic Surgeons score of 10; 1 paper provided a Japan SCORE of 9.4. Thirty-day mortality was 10% (10/104 patients, 7 of whom had acute TAAD). Causes of death were intraoperative rupture and tamponade (n = 2/104), late tamponade (n = 1/104), cardiac arrest (n = 1/104), type Ia endoleak (n = 1/104), sepsis (n = 1/104), multiple organ failure (n = 1/104), gastrointestinal bleeding (n = 1/104), severe aortic regurgitation and cardiac arrest due to the migration of the stent graft (n = 1/104) and unknown cause (n = 1/104). Mean duration of follow-up was 21 ± 21 months for 86 patients (10 patients died in-hospital and 8 case reports did not provide long-term outcomes). Ten out of 86 patients died during the follow-up period of non-aortic-related causes.

Procedural details, access route and endografts

All procedures but 2 (epidural anaesthesia) were performed with the patient under general anaesthesia with temporary pacing wires in the right ventricle. The most common access routes for endograft implantation were transfemoral (64/88; 73%), transapical (12/88; 14%) and combined transapical + transfemoral (6/104; 6%) (Fig. 2). One study with 16 cases reported unspecified transfemoral, transapical and transaxillary access routes. Some studies investigated in detail the anatomical characteristics of the dissected ascending aorta (data available for 50/104 cases): the mean distance between the primary entry tear and the right coronary artery or sinotubular junction was 35 ± 16 mm (range 7–61 mm); the mean distance between the primary entry tear and the innominate artery was 31 ± 17 mm (range 5–61 mm). A total of 114 stent grafts were used, including 12 custom-made devices, 65 off-the-shelf stent grafts and 7 modified stent grafts (Table 2). Data from 30 grafts were not available. The average length of the endovascular stent was 77 ± 29 mm (range 31–200 mm), and the mean diameter was 39 ± 5 mm (range 28–50 mm) in 90 reported cases. In 5 reports, the stent graft size was based on 5–20% oversize of the true lumen diameter [6, 10, 14, 25, 32]. In all cases, the proximal landing zone was in the proximal ascending aorta for entry tear exclusion, whereas the majority of the distal landing zones were proximal to the innominate artery origin. When the entry tear was near the innominate artery, in order to extend the distal landing zone, additional procedures were performed, including carotid-carotid bypasses or chimney grafts. Four studies did not report distal landing zone details (Table 2).

Figure 2:

Access routes for catheter-based repair of type A aortic dissection with the primary entry tear located in the ascending aorta. TA: transapical; TAX: transaxillary; TC: transcarotid; TF: transfemoral; TS: transseptal.

Table 2:

The details of the devices and anatomical data for the dissected ascending aorta

Author	Number of SG	SG type (CMD/OTS/modified on table)	Company	SG diameter (mm)	SG length (mm)	SG selection	Proximal position	Distal position	AA diameter (mm)	STJ diameter (mm)	Distal AA diameter (mm)	TL diameter (mm)	FL diameter (mm)	Distance between ET and CA/STJ (mm)	Distance between ET and IA (mm)
Li et al. [6]	15	OTS	Cook	39 ± 3.5	79 ± 2	Oversize of AA true lumen diameter proximal to entry tear: 20% in chronic, 10% in acute	DOCA	POIA	56 ± 7	50 ± 6	49 ± 10.2	P: 24 ± 7 D: 33 ± 5	P: 26 ± 8 D: 14 ± 6	50 ± 23	42 ± 18
Tsilimparis et al. [9]	16		Cook Zenith	45 ± 3 (range: 38–50)	73 ± 11 (range: 65–90)		DOCA
Grieshaber et al. [10]	6	CMD	Bolton		76	Based on aortic diameter at STJ	DOCA					19, 38, 34, 18, NA, NA	8, 4, 14, 45, NA, NA	36 ± 16 (range: 7–54)	21 ± 13 (range: 11–46)
Nienaber et al. [11]	12	OTS	Cook, Gore, Bolton, Optimed	34 ± 1 (range: 32–36)	72 ± 13 (range: 50–100)		DOCA
Roselli et al. [12]	14		Cook, Gore, Medtronic				DOCA		59 ± 10 (a) 47 ± 6 (c)	45 ± 10 (a) 39 ± 2 (c)	41 ± 6 (a) 47 ± 5 (c)	36 ± 5 (a) 36 ± 1 (c)		18 ± 3 (a) 51 ± 10 (c)	41 ± 20 (a) 23 ± 3 (c)
Bernardes et al. [13]	4	OTS	Cook, Gore, Medtronic	36 ± 3 (range: 32–40)	145 ± 64 (range: 80–200)		DOCA	POIA or zone 4	93 ± 15	31.3 ± 5
Ye et al. [14]	10	OTS	Medtronic, Cook, Ancure-II, Aegis	30–42	60–80 (100 in 1)	5–10% Oversize	DOCA	POIA or extent by prior LCCA-RCCA bypass						25–60	5–40
Wamala et al. [15]	2	Modified	Jotec	44, 40	75, 70		DOCA	POIA	41
Khoynezhad et al. [16]	2	OTS	Medtronic	38, 38	50, 50		DOCA	POIA
Murakami et al. [17]	1	OTS	Cook				DOCA	NA, LCCA treated with chimney graft + LCCA-RCCA bypass
Yang et al. [18]	1	CMD	Cook	38	77		DOCA	POIA
Khoynezhad et al. [19]	3	OTS	Medtronic	44 ± 3 (range: 40–46)	60 ± 20 (range: 40–60)		DOCA	POIA	56 ± 2
Vallabhajosyula et al. [20]	3	OTS	Cook	39 ± 1 (range: 38–40)	78 ± 2 (range: 77–81)		DOCA	POIA
Wilbring et al. [21]	1	OTS	Bolton	34	109		DOCA	POIA
Tauchi et al. [22]	1	OTS	Gore	28	33		DOCA	POIA
Luo et al. [23]	1	Modified	Medtronic	38	69		DOCA	POIA
Berfield et al. [24]	1	Modified	Cook	34	50		DOCA	POIA
Ronchey et al. [25]	5	OTS	Cook	40 ± 2 (range: 38–42)	57 ± 17 (range: 31–73)	At least 10% oversize in respect to the true lumen	DOCA	POIA or extent by LCCA-RCCA-RSA bypass	36 ± 3					30 ± 13 (range: 15–45)	To IA >5; to LSA: 55 ± 23
Shabaneh et al. [26]	1	OTS	Cook	40	81		DOCA	POIA
Pontes et al. [27]	1	CMD	NA	40	100		DOCA	POIA
Pinaud et al. [28]	2	OTS	Jotec				DOCA	POIA
McCallum et al. [29]	3	Modified	Medtronic	40	91 ± 41 (range: 44–115)		DOCA	POIA		33	42
Metcalfe et al. [30]	1	CMD	Cook	34	85		DOCA	POIA
Ruchat et al. [31]	1	OTS	EndoFit	44	70		DOCA	POIA					–
Palma et al. [32]	1	CMD	Braile Biomedical	38	155	20% oversize	DOCA	POIA				32	95
Senay et al. [33]	1	OTS	Medtronic		46		DOCA	POIA	62				43
Zimpfer et al. [34]	1	CMD	Jotec	46	85		DOCA	POIA
Zhang et al. [35]	1	OTS	Z-stent	38	70		DOCA	POIA	33
Ihnken et al. [36]	1	OTS	Gore	44	100		DOCA	POIA
Wang et al. [37]	1	CMD	Z-stent	38	70		DOCA	POIA
Dorros et al. [38]	1	OTS	Lacteba	35	35		DOCA	POIA
Total	114 SG/104 patients	OTS: 65/114 CMD: 12/114 modified: 7/114 NA: 30/114		39 ± 5 (28.5–50) (data available for 88 SG)	77 ± 29 (31–200) (data available for 97 SG)		DOCA: 104/104	POIA: 46/104 cases POIA or extent by epi-aortic bypasses: 19/104 NA: 39/104						35 ± 16 (7–61) (data available for 50/104 cases)	31 ± 17 (5–61) (data available for 50/104 cases)

Values are expressed as n (%) or mean ± SD.

a: acute; AA: ascending aorta; c: chronic; CA: coronary artery; CMD: custom-made device; D: distal; DOCA: distal to the origin of the coronary artery; ET: entry tear; FL: false lumen; IA: innominate artery; LCCA: left common carotid artery; LSA: left subclavian artery; NA: not available; OTS: off-the-shelf; P: proximal; POIA: proximal to the origin of the innominate artery; RCCA: right common carotid artery; RSA: right subclavian artery; SD: standard deviation; SG: stent graft; STJ: sinotubular junction; TL: true lumen.

Complications and reinterventions

Complications are listed in Table 3 and in Fig. 3 and include, among others, bailout conversion to open surgery (2/104 cases), endoleak (9/104 cases detected in-hospital and 4/86 detected during the follow-up period), endovascular reintervention (5/104 early and 1/86 late) and neurological complications (2 strokes during hospitalization and 3 during the follow-up of 86 cases). Type I endoleak was the most common complication: 7 cases were watched closely or the treatment was refused by the patient (1 case) [6, 12, 14, 20, 29, 33]. Newly detected diseases of the ascending aorta included dissection in the arch (n = 1/104), recurrence of TAAD (n = 2/83), pseudoaneurysm (n = 2), a new penetrating ulcer in the ascending aorta with haematoma (n = 1/83) and an aneurysm at the distal landing zone (n = 1/83). The patient with new aortic arch dissection was treated with a branched endograft [19]. Four patients who had retrograde dissection and ventricular pseudoaneurysm, recurrence of dissection and a new penetrating ulcer had open surgery [6, 13, 14]. Two more underwent endovascular repairs [6, 14]. Three patients had coronary complications: 2 patients were treated with coronary percutaneous procedures [6, 10] whereas 1 patient with partial left ostium obstruction required emergency open surgery on cardiopulmonary bypass [12]. Nevertheless, he died 6 days later of stroke and multiple organ failure. Other complications included 1 patient with a retained delivery system requiring urgent conversion to open surgery [12]; 1 case of late severe aortic valve insufficiency requiring a Bentall procedure 21 months later [35]; and 2 cases of intraprocedural severe aortic valve regurgitation due to leaflet entrapment requiring balloon-repositioning of the endograft [19, 32].

Figure 3:

Hospital and late outcomes of catheter-based repair of type A aortic dissection with the primary entry tear located in the ascending aorta.

Table 3:

Summary of 30-day and long-term complications following endovascular repair of type A aortic dissections

Complication	Overall	30-Day	Follow-up	Reintervention
Endoleak	13	9	4	In-hospital endovascular repair: 5 [12, 13, 26] Late open repair: 1 [12] Watch closely: 6 [6, 14, 20, 29, 33] Patient refused treatment: 1 [12]
Stroke	5	2	3	Not required [12, 14, 20]
New ascending aorta or arch disease; progression of the dissection	7	1	6	Endovascular repair 3 and 8 months later: 2 [6, 14] Open root surgery 29 months later: 1 [6] Open ascending aorta surgery 2.2 and 1.7 months later: 2 [13] Open ascending aorta and arch surgery with frozen elephant trunk procedure 2 months later: 1 [19] Watch closely: 1 [18]
Left coronary ostium obstruction	1	1	0	Conversion to open surgery (intraprocedural) with circulatory arrest and ascending aorta replacement: 1 [12]
Severe aortic valve regurgitation	3	2	1	Intraprocedural balloon repositioning of the endograft: 2 [19, 32] Open surgery for Bentall operation 21 months later: 1 [35]
Coronary stent thrombosis	1	1	0	Coronary stenting (intraprocedural) and ECMO support for 3 days [10]
Myocardial ischaemia	1	1	0	Coronary stenting (intraprocedural) [6]
Retained endograft delivery system	1	1	0	Conversion to open surgery (intraprocedural) with circulatory arrest, removal of the delivery system and direct stent graft fixation [12]
Pericardial effusion	1	1	0	Conservative treatment [6]
Pulmonary embolism	1	1	0	Conservative treatment [13]

ECMO: extracorporeal mechanical oxygenation.

DISCUSSION

Acute TAAD requires emergency surgical correction for ascending aorta and partial or total arch replacement. Despite improvements in surgical and anaesthesiological techniques, hospital deaths and the neurological complication rate remain high in large registries [1–4, 39]. In older patients or in patients with severe comorbidities, surgery for TAAD is associated with relatively higher intraoperative mortality with 30-day survival achieving 60% [1–4, 40]. Since the first reported case of endovascular repair of TAAD by Dorros et al. [38] in 2000, the endovascular treatment has slowly became a potential alternative to open surgery in selected high-risk or inoperable patients.

In this review, several retrospective studies and case series reported acceptable results for patients receiving endovascular treatment for TAAD with the entry tear in the ascending aorta. Grieshaber et al. [10] reported 6 patients (EuroSCORE II 34%) with acute TAAD treated endovascularly: during a mean follow-up time of 23 months, no deaths or complications occurred. Another retrospective analysis of 15 patients (American Society of Anesthesiologists: IV or greater) reported by Li et al. [6] also showed 100% survival after a mean follow-up time of 72 months. In 24 case reports, 2 out of 29 patients (7%) died in-hospital. Considering the indication, in 5 cases (5%) the patients had already been operated on for aortic root or ascending aorta replacement (in 3 cases because of a TAAD) and the treatment was directed towards the on-going distal dissection or the pseudoaneurysm of the distal ascending aorta. It is important to mention that, from a procedural point of view, the placement of an endograft in a previously implanted surgical prosthesis is simpler and safer compared to placing the same endograft in a freshly dissected aorta with fragile wall tissue.

These preliminary results suggest that endovascular treatment for TAAD with the entry tear in the ascending aorta has acceptable early mortality (10%) and stroke (2%) rates. An interesting finding is that 7 of 10 deceased patients, including the 2 intraoperative deaths, suffered from acute TAAD, confirming that endovascular treatment remains a high-risk procedure in patients with acute dissection because of the more fragile aortic wall tissue and the critical clinical condition of the patient.

If we consider the complications, the reintervention rate (5% in-hospital and 8% late) is significant; the major causes were endoleaks and new ascending aorta diseases. Six cases were converted to open surgery either during the procedure or within 3 months, suggesting that patient selection and stent graft selection is crucial. Moreover, in 1 case, extracorporeal mechanical oxygenation support was required for 3 days after a concomitant coronary intervention and myocardial ischaemia. Therefore, in order to be ready for urgent cardiopulmonary bypass or extracorporeal mechanical oxygenation instauration in such high-risk patients requiring high-risk endovascular interventions, a surgical femoral vascular access should always be provided and a perfusion team together with a cardiac surgeon should be on standby in the hybrid or operating room.

Different models of and strategies related to the appropriate selection of stent grafts were applied in the absence of a specifically designed prosthesis. The majority of the stent grafts were ‘off-the-shelf’ endografts; some procedures [9, 25, 30] were performed with custom-made stents (Zenith Ascend TAA Endovascular Graft; William Cook Europe, Bjaeverskov, Denmark); in other cases, the surgeons modified standard endoprostheses. Currently available stent grafts do not fully address the unique features of the anatomy of the ascending aorta. Moreover, the development of specifically designed devices is a challenge not only because of the complex anatomy and physiology but also because of the absence of healthy segments for safe fixation of the stent graft in the dissected tissue. Also, with the material currently available, the anatomy of the dissected aorta is crucial: strict patient selection with proximal or distal landing zone length equal to or greater than 20 mm and an entry tear location in the middle part of the ascending aorta will improve early and later outcomes as suggested by Li et al. [6].

Several access routes have been proposed for endovascular treatment of TAAD. The initial access reported by Dorros et al. was transseptal. In our review, the transfemoral route was the most commonly used, and 18 cases were performed through a transapical or combined transapical and transfemoral approach. Compared with the transseptal route, transapical access provides a short working distance to the ascending aorta, giving the operator a significantly higher degree of stability and control during the deployment. The transapical approach also prevents the guidewire from entering into the false lumen and facilitates pericardial drainage if required.

In the treatment of aortic dissection, time is crucial. To buy time for standard reconstructive open surgery, Tauchi et al. [22] reported a case successfully managed by rescue stent grafting of the ascending aorta from the carotid artery for life-threatening myocardial ischaemia following acute aortic dissection.

Aortic regurgitation in TAAD is commonly related to the dilation of the aortic root and annulus, tearing of the annulus or valve cusps, downwards displacement of 1 cusp below the line of the valve closure, loss of support of the cusp and physical interference in the closure of the aortic valve by an intimal flap [7]. In the reported papers, none of the patients had severe aortic valve regurgitation at the time of the procedure; therefore, information on the impact of catheter-based treatments on this clinical issue is missing. Nevertheless, Wilbring et al. [21] reported the complete concomitant endovascular treatment of acute TAAD and severe aortic stenosis by performing, during the same procedure, a transfemoral transcatheter aortic valve replacement and thoracic endovascular aortic repair for the ascending aorta. This approach suggests an endovascular strategy for complex aortic root disease, which requires extension of the proximal landing zone to the aortic root [41]. Composite endovascular devices may further facilitate this approach for treatment of TAAD in the near future, but the development of such devices is still embryonic [42].

Limitations

Despite the good procedural success rate and despite the reports concluding that the endovascular treatment is a minimally invasive effective method to treat TAAD in high-risk patients, the anatomical and functional properties of the dissected ascending aorta (separation of the aortic wall layers, torn intimal, the relationship between the primary tear and the coronary ostia or supra-aortic branches) still limit the suitability of this portion of the aorta for endovascular treatments. Nevertheless, the presence of the primary tear in this region can lead to full recovery of the aorta after endovascular treatment if endoleak is not detected. However, it is worth mentioning that significant anatomical and physiological variability exists among patients, and TAAD represents a complex evolving disease. Careful sample selection is required before drawing any conclusions from the presented data [43]. The main limit of the present study is that it is a review paper collecting data from published case series and case reports and there is a risk of high publication bias.

CONCLUSION

The available literature on endovascular treatment of the dissected ascending aorta with the primary entry tear in the ascending aorta still consists of case reports and small case series with a limited number of studies describing the long-term outcomes. Currently, use of the endovascular treatment can only be considered in selected high-risk patients who are not good candidates for standard surgery. The future development of specifically designed devices will further improve the outcome of this new technique.

Conflict of interest: Dr Maisano received grant and/or research support from Abbott, Medtronic, Edwards Lifesciences, Biotronik, Boston Scientific Corporation, NVT, Terumo. He received consulting fees or honoraria from Abbott, Medtronic, Edwards Lifesciences, Perifect, Xeltis, Transseptal solutions, Cardiovalve, Magenta. He has Royalty Income/IP Rights from: Edwards Lifesciences (FMR surgical annuloplasty). He is shareholder (incl. stock options) of: Cardiovalve, Magenta, Transseptal solutions, 4Tech, Perifect, Coregard. Dr Ferrari received Educational grants from Edwards Lifesciences and Medtronic and is Consultant for Edwards Lifesciences.

Author contributions

Changtian Wang: Data curation; Formal analysis; Methodology; Validation; Writing—original draft. Ludwig Karl von Segesser: Conceptualization; Data curation; Investigation; Methodology; Writing—review & editing. Francesco Maisano: Supervision; Validation. Enrico Ferrari: Data curation; Formal analysis; Validation; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Martin Czerny, Daniel-Sebastian Dohle, Sven Peterss and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

Abbreviation

 
• TAAD

  Type A aortic dissection