Treatment of acute type A aortic dissection with the Ascyrus Medical Dissection Stent in a consecutive series of 57 cases

Abstract

Open in new tabDownload slide

OBJECTIVES

The Ascyrus Medical Dissection Stent (AMDS) has been recently introduced as an alternative for total arch replacement in acute aortic dissection type A (AADA). The aim of this study was to evaluate the postoperative outcomes after AMDS treatment in a large contemporary cohort of AADA patients.

METHODS

Data acquisition was performed retrospectively at 2 German aortic centres between 2020 and 2022 and comprised the perioperative parameters and postoperative results of all AADA patients. All patients treated with the AMDS for AADA were included in the study. The primary end point was in-hospital mortality. Secondary end points were defined as early postoperative and AMDS-related complications.

RESULTS

Fifty-seven AADA patients treated by AMDS were included in the study group. The mean age was 64.6 ± 10.8 years and 59.7% (n = 34) were males. The actual in-hospital mortality was considerably lower than the predicted mortality risk by the German registry for acute aortic dissection type A score (16% vs 22%). The median ICU and in-hospital stay were 5 (interquartile range: 3–13) and 12 (interquartile range: 10–22) days, respectively. Postoperative complications comprised acute renal insufficiency (37%) with need for temporary (16%) or permanent dialysis (5%), delirium (26%), re-exploration for bleeding (14%), tracheostomy (14%) and new stroke (4%). A new AMDS-related complication (central stent collapse) was observed in 9% (n = 5) by postoperative computed tomography and chest X-ray. The incidence of complete central AMDS collapse did not impact 30-day mortality.

CONCLUSIONS

The AMDS may be successfully used in AADA with acceptable 30-day mortality in accordance with the German registry for acute aortic dissection type A score. However, careful preoperative evaluation of the patient’s individual aortic anatomy regarding potential contraindications and proper device implantation are strongly recommended to avoid complete central AMDS collapse.

INTRODUCTION

Acute aortic dissection type A (AADA) remains one of the most dreadful clinical emergencies with rapidly increasing mortality until diagnosis and immediate surgical treatment [1, 2]. Most often AADA extends from the proximal thoracic aorta into the aortic arch and further downstream the aorta. In these scenarios, the supra-aortic branches are likely to be affected by dissection, potentially resulting in cerebral malperfusion with clinical neurology or even coma. Traditionally, total arch replacement with complete reconstruction of the supra-aortic branches or extra-anatomic bypasses (as a bail-out strategy) are required to save the patient’s life [3]. However, the associated postoperative mortality after total arch repair in AADA patients remains high with reported mortality rates of 17–26% [4, 5].

More recently, the Ascyrus Medical Dissection Stent (AMDS) has been introduced as a new tool for AADA treatment, potentially avoiding prolonged cardiopulmonary bypass times and complex aortic arch replacement in well suited patients [6, 7]. Early results with the use of the AMDS are promising but are still limited due to small cases numbers [8]. Therefore, the aim of this clinical study was to evaluate the postoperative outcomes of AADA patients treated with the AMDS hybrid prosthesis in a large consecutive series of 2 high-volume aortic centres.

MATERIALS AND METHODS

Data acquisition was performed retrospectively at 2 German aortic centres between January 2020 and June 2022 and comprised the perioperative parameters and postoperative results of all AADA patients. All patients treated with the AMDS hybrid prosthesis for AADA were included in the study group. Preoperative baseline characteristics, intraoperative data and postoperative outcome variables were collected, anonymized and entered into a standardized database for statistical analysis. Prediction of 30-day mortality for comparison with the observed 30-day mortality of the study group was performed by using the web-based calculators of the EuroScore II and the German registry for acute aortic dissection type A (GERAADA) score, respectively [9, 10].

Ethics statement

The local Ethics Committees of the University of Cologne (No. 20-1212) and the University of Frankfurt (No. 19-378) both approved the study. Individual patient consent was waived due to the retrospective design of the study.

Device description and surgical technique

The AMDS (Artivion, Kennesaw, GA, USA) consists of a PTFE felt graft (cuff) and an uncovered nitinol wire braided stent to allow for a proximal arch anastomosis (zone 0) and true lumen stabilization with potential induction of positive aortic remodelling of the aortic arch and descending aorta [11]. The AMDS is available in 2 different shapes of the nitinol stent (straight and tapered) and 2 different felt cuff sizes (24 vs 32 mm), respectively. The required sizing is performed preoperatively using a multiplanar computed tomography (CT) scan with 2 aortic landmarks: zone 1 in the aortic arch and the descending aorta at the tracheal bifurcation (zone 4).

In brief, the surgical technique comprises the use of cardiopulmonary bypass and ascending aortic replacement, with or without root replacement (depending on the extent of AADA). Once the desired level of hypothermia is achieved, the arch is opened and inspected under additional cerebral protection by antegrade selective perfusion. In the absence of a re-entry tear or large aneurysm in the arch or descending aorta (contraindication), the AMDS is inserted into the arch. After four-fold fixation of the felt cuff to the aorta (zone 0; at least 10–20 mm prior to the inomminate artery), the stent is released from the guiding device and the felt cuff is anastomosed to the proximal aortic arch. Afterwards, the procedure is completed by an anastomosis of the ascending prosthetic graft to the AMDS felt cuff with a running suture (Fig. 1A). The AMDS device implantation is shown in Video 1.

Definitions and statistical analysis

The 30-day mortality was defined as any death within the first 30 days after surgery. The term postoperative was defined as the time interval after AMDS surgery until the last contact with the patient during follow-up. End-organ malperfusion was assessed clinically (evident clinical symptoms) and radiographically by CT imaging (pre- and postoperatively). Malperfusion was defined as the loss of adequate blood supply to a vital organ resulting from arterial obstruction secondary to the dissection. The Tear, Entry site, Malperfusion (TEM) classification was used to classify the type, entry site and organ malperfusion (with or without clinical symptoms) [12]. Renal insufficiency was defined as a rise in serum creatinine levels of more than 0.3 mg/dl (26.5 μmol/l) over 48 h or a rise in serum creatinine 1.5 times higher than average levels newly developed reduction of urine at <0.5 ml/kg/h over 6 h. Preoperative neurologic deficits were defined as any neurologic damage present prior to surgery. Postoperative neurologic deficits were defined as any neurologic damage—temporary or permanent—diagnosed after surgery.

The primary end point was defined as 30-day mortality. Secondary end points were defined as early clinically relevant postoperative complications such as respiratory (reintubation), renal (acute kidney failure with or without need for dialysis) insufficiency and new temporary (e.g. delirium) or permanent (e.g. stroke and paraplegia) neurology and AMDS-related complications. An AMDS-related complication was diagnosed either by CT scan or (if unavailable) by postoperative chest X-ray.

Categorical variables were reported using absolute and relative frequencies. Continuous variables were expressed by mean ± standard deviation or as median with interquartile range [interquartile range (IQR); 25th to 75th percentile]. Tests between groups were performed using chi-squared tests for categorical variables and t-tests for continuous variables. Since this is a study of exploratory character, P-values were not adjusted for multiplicity and had to be interpreted descriptively. The impact of the available pre-, intra- and postoperative variables (n = 95) on the 30-day mortality was analysed using a multivariable regression model. A binary logistic regression model was constructed with the available variables and the risk factors with a P-value of <0.1 from univariable analysis (Supplementary Material, Table S1). A two-tailed P < 0.05 or less was considered to be statistically significant. Survival estimation was performed by the Kaplan–Meier method. Statistical analyses were performed using SPSS software (version 27.0, IBM, Armonk, NY, USA) for data analysis and visualization.

Subgroup analysis

A subgroup analysis was performed for the comparison of patients with or without AMDS-related complications. All patients with at least 1 postoperative CT scan before hospital discharge were included the subgroup analysis, while 8 patients without postoperative CT scans were excluded. Respective measurements of the inner and outer length of the ascending aortic graft (distance from the sinotubular junction to the AMDS felt cuff) and the AMDS stent portion were performed by a radiologist blinded to the study protocol and results (Fig. 1B).

RESULTS

Preoperative baseline characteristics and dissection-specific data

A total of 57 (30%) out of 196 AADA patients from both centres met the inclusion criteria and were included in the study group. The mean age was 64.6 ± 10.8 years, and 60% (n = 34) were males.

Preoperatively, relevant comorbidities of the study patients comprised arterial hypertension (n = 44; 77%), renal insufficiency (n = 11; 16%), hyperlipoproteinaemia (n = 8; 14%), coronary heart disease (n = 6; 11%), COPD (n = 4; 7%) and diabetes mellitus (n = 3; 5%). On admission, dissection-specific complications were noted as regurgitation of the aortic valve (n = 35; 61%), pericardial tamponade (n = 11; 19%), neurologic deficit (hemiparesis 11% and paraparesis 9%), need for ventilation (n = 7; 12%), need for inotropes (n = 7; 12%) and previous cardiopulmonary resuscitation (n = 4; 7%). The details of relevant preoperative baseline characteristics are depicted in Table 1.

The study patients suffered from either hyper-acute (n = 54; 95%), acute (n = 2; 4%) or subacute (n = 1; 2%) aortic dissection, including 55 type A (96%) and 2 non-A-non-B (4%) aortic dissections. Aortic dissection extension ended at the level of the descending aorta (n = 7), abdominal aorta (n = 12) and below the iliac bifurcation (n = 35) in 12%, 22% and 61%, respectively. The supra-aortic vessels were involved by dissection in 55 cases (96%). Preoperatively, organ malperfusion was classified with regard to the TEM classification as coronary (M1: n = 9; 16%), supra-aortic (M2: n = 16; 28%) and spinal/viszeral/iliac (M3: n = 23; 40%). Clinical symptoms of malperfusion were evident in 41 patients (72%). The respective TEM classification is shown in Table 1.

Surgical risk predictions for 30-day mortality were calculated with the web-based GERAADA score (21.7 ± 14.1%) and the EuroScore II (15.7 ± 11.5%) calculators (Table 1).

Intraoperative data

Cannulation for cardiopulmonary bypass (CPB) was performed by right axillary (97%) or direct aortic with additional left axillary cannulation (3%). Antegrade selective cerebral perfusion was used in all cases, either uni- (n = 23; 40%) or bilaterally (n = 34; 60%), with application of moderate hypothermia (28.0±1.1°C; mean body core temperature). All study patients received ascending aortic replacement in combination with the AMDS. In addition, isolated aortic valve replacement was necessary in 16 (28%) cases, while root procedures were performed as a Bentall or valve sparing operation in 10 (18%) and 1 (2%) patients, respectively. Concomitant CABG (n = 12; 21%) was performed if reimplantation of both coronary buttons during the Bentall procedure was not feasible (calcified or dissected) or if CPB weaning was unsuccessful due to low out-put with signs of left or right ventricular failure and a known history of coronary artery disease. Mitral valve repair was performed in a single case (2%). The details of intraoperative data are given in Table 2.

Postoperative outcomes

The actual in-hospital mortality was 16% for the study cohort. The median ICU and in-hospital stay were 5 (IQR: 3–13) and 12 (IQR: 10–22) days, respectively. The mean follow-up time was 4.3 ± 5.6 months. The estimated survival was 83% after 21 months (Fig. 2). Postoperative complications included acute renal insufficiency (37%) with need for temporary (16%) and permanent (5%) dialysis, delirium (26%), re-exploration for bleeding (14%), tracheostomy (14%) and new stroke (4%). The postoperative outcomes are shown in detail in Table 3.

A new undescribed AMDS-related complication (central collapse) was observed in 9% (n = 5) by postoperative computed tomography and chest X-ray before discharge (Table 3): the proximal and distal AMDS portions were inflated while a complete central stent collapse was evident in each of the 5 respective patients (Fig. 3).

Risk factor analysis

The available perioperative variables (n = 95) from the database, including the AMDS collapse, were used for risk factor analysis for 30-day mortality of the study cohort. Univariable analysis identified preoperative ventilation (P = 0.004) as a significant predictor for 30-day mortality (Supplementary Material, Table S1). No independent risk factors for in-hospital mortality could be identified by multivariable analysis.

Subgroup analysis

Two subgroups with available postoperative CT scans were compared: AMDS collapse (n = 5; 9%) versus patients without observed AMDS-associated complications (n = 44; 77%). Postoperative CT scans were unavailable in 8 patients.

In the 5 patients with a central AMDS collapse (9%), the inner ascending aortic graft length was found to be significantly shorter than in patients without collapse (30.0 ± 5.9 vs 39.6 ± 10.9 mm; P = 0.029). There also was a trend towards shorter outer ascending graft length in the AMDS collapse subgroup (48.5 ± 9.5 vs 55.7 ± 14.7 mm; P = 0.066). However, total length of the stent portion of the AMDS did not significantly differ between patients with and without central AMDS collapse (P = 0.39) (Table 4).

DISCUSSION

Open surgery remains the gold standard of treatment AADA patients who will otherwise face a mortality of 1-2% per hour, leading up to 50% within the first 48 h. The treatment of AADA is complex and may become very challenging if the aortic arch is affected by the dissection (DeBakey type I aortic dissection) with reported in-hospital mortality rates of up to 26%—while in the in the presence of end-organ malperfusion early postoperative mortality may even increase up to 40% [5, 11]. In this scenario, the frozen elephant trunk (FET) procedure has been suggested for total arch replacement by renowned aortic surgeons, since this procedure allows for coverage of potential re-entries and may facilitate positive aortic remodelling in the downstream aorta [13, 14]. However, this procedure usually requires a dedicated aortic team with experienced surgeons to achieve exceptional results and low postoperative mortality rates [15].

The hemiarch procedure—as an alternative approach to treat AADA—may be performed by (less experienced) surgeons with shorter cardiopulmonary bypass times and low postoperative in-hospital mortality rates [15–17]. However, this technique avoids total arch repair and bares a considerable risk for distal anastomotic new entry (DANE) with patent false lumen perfusion—ultimately leading to aortic growth, end-organ malperfusion and worse long-term survival [18, 19]. Recently, the AMDS hybrid prothesis was introduced as a new tool in the surgeon’s armamentarium to treat AADA without the need for total arch repair [7].

The dissected aorta repair through stent (DARTS) trial reported an excellent in-hospital and 1-year mortality after AMDS treatment of 13% and 19.6%, respectively [11]. Most recently, Montagner et al. [20] published ‘real-world’ data from a large single centre series of 100 AMDS cases with a reported 30-day mortality of 18%. Thus, the reported 30-day mortality of the underlying contemporary surgical series of 16% in 57 AADA cases is in line with the current literature, especially since >70% of the study group presented with preoperative end-organ malperfusion. Interestingly, the GERAADA score—specifically designed for AADA risk prediction—overestimated the in-hospital mortality of the study cohort (21.7 ± 14.1%) while the prediction by the EuroScore II (15.7 ± 11.5%) was more accurate in this specific series of AADA patients.

Postoperative complications after surgery for extensive AADA usually affect postoperative kidney function due to preoperative malperfusion and CPB duration [21]. In this series, acute renal insufficiency was the most frequent postoperative complication (37%) with need for temporary or permanent dialysis in almost one-quarter (23%). This finding is not surprising since 82% of study patients had a dissection extension into the abdominal aorta with 23 patients suffering from M3 malperfusion due to TEM classification.

The second highest incidence of postoperative complications were noted to be temporary delirium (26%). This finding is neither surprising since the prevalence of preoperative cerebral malperfusion (M2) in the study group was equally high (28%). Liu et al. [4] reported the incidence of postoperative delirium after AADA surgery to be up to 34% in their series of 100 AADA cases, identifying cerebrovascular disease, operation and CPB duration, intubation time and postoperative hypoxia as significant predictors. In comparison, the CPB time, operation time and ICU stay in our study were considerably shorter which might explain the lower incidence of postoperative delirium.

In this study series, permanent neurologic complications, by means of new postoperative stroke (4%) and spinal cord injury (0%), were noted to be very low. Supra-aortic involvement by the dissection—with resulting cerebral malperfusion—has been suggested to potentially result in higher stroke rates ranging between 11% and 46.7% [11]. Bozso and Montagner both reported equally low stroke rates of 6.3% and 8% in their respective AMDS series, thereby advocating the usefulness of the new hybrid prosthesis for AADA treatment [11, 20]. In 2016, Leontyev et al. [22] reported a similar stroke rate (7.7%) but higher incidences of postoperative spinal cord injury (7.7%) in a large multicentre series of 509 patients treated by FET—including 350 AADA cases. In their meta-analysis, Mousavizadeh et al. [23] reported excellent results with the FET technique after zone proximalization for in-hospital mortality, cardiovascular events and paraplegia of 7%, 6% and 3%, respectively. Most recently, Mehdiani et al. [24] also reported exceptional results with the AMDS prothesis in 8 AADA patients without any postoperative neurologic complications despite the need for extra-anatomic reconstruction.

In 2017, Rylski et al. [18] reported a DANE incidence of up to 70% after limited aortic repair (e.g. hemiarch repair). In addition, the incidence of DANE was identified as an independent risk factor for distal aortic events during postoperative follow-up—with worse long-term survival compared to patients without DANE by over 10% at 5 years and over 30% at 10 years, respectively [19]. In this study, the early incidence of postoperative DANE was 0% which is in line with other AMDS clinical studies [8, 11, 24]. These results as well as the reported high incidences of positive aortic remodelling after AMDS treatment during early follow-up are very promising; however, mandatory mid- and long-term follow-up after AMDS treatment is still lacking.

Postoperative CT imaging revealed complete central AMDS collapse as a new and previously unknown complication after AMDS treatment in 5 (9%) of the study patients. Remarkably, the proximal and distal stent portions remained inflated, seemingly unaffected by the local and centralized AMDS collapse (Fig. 3). Theoretically, a tubular body, like the AMDS hybrid prosthesis, can only bend if the inner curvature is compressed and the outer curvature is stretched. Stretching of the uncovered AMDS stent may lead to a decrease in diameter, resulting in flat angles of the stent braid and subsequently a reduction of flexibility.

In this context, the authors hypothesize that a complete central AMDS collapse is more likely to occur if 1 or a combination of the following conditions is present:

1. Unfavourable anatomy:

   • Gothic arch and

   • Aortic kinking;

2. Device application:

   • AMDS oversizing,

   • Suboptimal aortic transection and

   • Increased proximal tension;

3. Aortic elongation during reperfusion.

In contrast to the FET procedure, the arch is not completely resected during AMDS treatment. Therefore, the aneurysmatic arch disease has already been acknowledged as a contraindication for AMDS treatment. However, the presence of a gothic arch or aortic kinking (e.g. in the proximal descending aorta) has not been considered disadvantageous for AMDS treatment so far. We believe these specific anatomic configurations result in an acute angle that might favour a central stent collapse once the ascending graft is anastomosed to the AMDS felt cuff.

The correct application of the AMDS hybrid prosthesis might also be in jeopardy if the current recommendations by the distributing company on implantation and sizing are ignored. As a matter of fact, oversizing remains a serious issue since there are only 2 available cuff sizes. Due to the authors’ own experiences, aortic surgeons are more reluctant to use the 24 mm cuff to avoid potential aortic stenosis despite a previously measured aortic diameter of 25–36 mm in zone 4, indicating the recommended use of the smaller device by the company. Another aspect to be considered is aortic transection—recommended being performed at least 10 mm proximal to the innominate artery. In general, the proximal arch is resected transversely towards the inner arch’s curvature during hemiarch repair. In addition, the ascending aortic graft is shortened and resected correspondingly, with a shorter portion at the inner arch curvature, to avoid the potential kinking of an elongated ascending graft. However, this practice increases the tension at the distal anastomosis of the ascending graft, thereby pulling the aortic arch prosthesis towards the sinotubular junction. Due to the configuration of a FET hybrid prosthesis with a Teflon covered stent portion, this manoeuvre is without consequence; however, the AMDS is fixed in the descending aorta after deployment and the applied tension towards the sinotubular junction (ultimately) might lead to increased stretching of the bare stent with resulting central collapse (Fig. 4). This theory is supported by the measurements of the ascending grafts: patients with central AMDS collapse had significantly shorter ascending grafts (Table 4).

Finally, after finishing the distal ascending anastomosis pulsatile blood flow is restored and aortic reshaping, with potential elongation of the pressurised native aorta may occur. Although the measurements performed in this study did not show a significant difference in elongation of the AMDS between groups, it seems likely that in combination with increased proximal tension and/or an unfavourable anatomy, even slight aortic reshaping (with elongation) might be the icing on the cake with a resulting central AMDS collapse.

Tsagakis et al. [25] reported on the potential benefits of intraoperative angioscopy of the aorta to exclude re-entries of the downstream aorta and to gain landing zone control before FET deployment, which could also be utilised for verification of adequate AMDS positioning and stent inflation. However, once the distal ascending anastomosis is completed and pulsatile blood flow has been restored, an acute AMDS collapse cannot be visualized without diagnostic imaging or angiography—which is seldom available in a conventional OR suite.

Therefore, the authors of this study strongly recommend (i) preventing increased tension towards the sinotubular junction—by leaving a longer ascending aortic graft and less radical aortic resection of the inner arch curvature—and (ii) avoiding AMDS oversizing, as well as acknowledging (iii) gothic arch anatomy and aortic kinking as a potential contraindication for AMDS treatment to reduce the likelihood of complete central AMDS collapse occurrence. In addition, in the presence of an AMDS collapse medical anticoagulation therapy is advised. Finally, continuous surveillance with routine CT imaging is mandatory for verification of an optimal postoperative result and providing long-term follow-up after AMDS treatment. However, more clinical studies—with larger patient cohorts are required to discover the potential causes of the reported central stent collapse and to elucidate the clinical impact of this new AMDS complication in the future.

Limitations

The retrospective study design—with early follow-up and limited secondary data availability—is acknowledged by the authors as a major weakness of the study. However, this descriptive clinical analysis discovered a new and previous unknown complication of the AMDS hybrid prosthesis. The authors strongly recommend future experimental and clinical studies with larger AMDS patient cohorts to identify potential risk factors of the collapse and to investigate on the potential benefits (e.g. positive aortic remodelling) during follow-up of this new and promising device.

CONCLUSIONS

The AMDS may be successfully used in AADA with acceptable 30-day mortality in accordance with the GERAADA score. However, careful preoperative evaluation of the patient’s individual aortic anatomy regarding potential contraindications, such as gothic arch anatomy, and proper device implantation with decreased proximal tension on the AMDS device are strongly recommended to avoid complete central AMDS collapse.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

Conflict of interest: The authors declare that there are no potential conflicts of interest regarding this particular study.

Data availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Author contributions

Maximilian Luehr: Conceptualization; Formal analysis; Investigation; Methodology; Project administration; Supervision; Validation; Writing—original draft. Christopher Gaisendrees: Conceptualization; Data curation; Formal analysis; Investigation; Resources; Validation; Writing—original draft. Abdul Kadir Yilmaz: Data curation; Formal analysis; Investigation. Leila Winderl: Data curation; Formal analysis; Investigation; Resources. Georg Schlachtenberger: Conceptualization; Data curation; Formal analysis; Software; Validation. Arnaud Van Linden: Supervision; Writing—review & editing. Thorsten Wahlers: Supervision; Writing—review & editing. Thomas Walther: Supervision; Writing—review & editing. Tomas Holubec: Conceptualization; Methodology; Project administration; Supervision; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Roman Gottardi, Davide Pacini and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

Presented at the 36th Annual Meeting of the European Association of Cardio-Thoracic Surgery (EACTS), Milano, Italy, 5–8 October 2022.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• AADA

  Acute aortic dissection type A

 
• AMDS

  Ascyrus Medical Dissection Stent

 
• CPB

  Cardiopulmonary bypass

 
• CT

  Computed tomography

 
• DANE

  Distal anastomotic new entry

 
• FET

  Frozen elephant trunk

 
• GERAADA

  German registry for acute aortic dissection type A

 
• IQR

  Interquartile range

Author notes