Long-term Outcomes of Intervention between Open Repair and Endovascular Aortic Repair for Descending Aortic Syndrome: A Propensity-Matched Analysis

Abstract


Background
The treatment of descending thoracic aortic syndromes is challenging and depends on its pathologies.
Since the introduction of thoracic endovascular aortic repair (TEVAR), many reports have demonstrated that it is a safe and feasible alternative to the conventional open repair [1][2][3].Typically, the success of this procedure is dictated by its favorable outcomes and ease of use [4].Although the use of TEVAR has rapidly increased due to improved perioperative morbidity rates, signi cant postoperative complications associated with TEVAR, such as endoleak, stent-graft migration, retrograde type aortic dissection, newonset dissection, and stent-graft infection, contribute to relatively poor results of TEVAR [3,5,6].In early clinical results with open repair versus TEVAR covered in previous reports [3,6,7], there are no long-term data comparing two procedures.In particular, the durability and long-term complication rates of open repair and TEVAR have not yet been determined.In this study, the primary end points were in-hospital mortality, and short-term complications.The secondary end points were long-term mortality and reintervention rates.
The purpose of this study was to compare long-term outcomes and reintervention rates between open repair and TEVAR in patients with descending aortic syndromes.To neutralize the effects of confounding independent variables such as unbalanced numbers (45:91) and age discrepancy, a propensity matched subsample of patients was created for an adequately powered analysis.

Study population
Between January 2002 and December 2017, 512 patients were diagnosed with descending thoracic aortic syndromes including dissection, pseudoaneurysm and penetrating atherosclerotic ulcer (PAU) at Kyungpook National University.The study design owchart is demonstrated in Fig. 1. 230 patients underwent surgery.Among them, 136 patients were included for the study.45 patients underwent open repair, and 91 underwent TEVAR.We retrospectively reviewed the medical records and undertook telephonic clinical assessments of these patients.

De nitions
Acute dissections was de ned when the occurrence developed in 14 days from the rst symptom.Complicated aortic syndromes were de ned as the presence of one or more of the following conditions: aortic maximum size > 5.5 cm; resistant hypertension despite adequate medical therapy; recurrent or refractory pain; impending rupture; rupture with end-organ malperfusion; and extension of dissection.
Aortic reintervention was de ned as the need for any surgical or endovascular interventions following the initial procedure during follow-up.In the TEVAR group, an endoleak was de ned as radiological evidence of blood ow outside the stent -graft according to published guidelines [8].

Operative strategies
The treatment modality was decided collaboratively by the cardiologist and cardiac surgeons who were involved in the patients' care; decision was based on the patients' co-morbidities, functional status, anatomical feature of the lesion, and the appropriate of vascular access [9].

Open aortic repair
Open aortic repair was generally performed via left thoracotomy or median sternotomy.All procedures employed with sequential clamping to minimize ischemic times.In case of left atrial-left femoral artery partial bypass (LHB), blood was drained from the left atrium via the inferior pulmonary vein and returned through the femoral artery.In case of circulatory arrest, bypass was initiated via the femoral artery and vein or the ascending aorta and right atrium.

TEVAR
Preoperative CT scans were reviewed for the preoperative assessment of access routes for the feasibility of TEVAR and to measure the bilateral vertebral artery for assessing the subclavian steal syndrome with a policy of selective subclavian artery revascularization.All TEVAR procedures were performed via the transfemoral approach under general anesthesia.Perioperative anticoagulation with heparin was prescribed at a dose of 3000-5000 units.The proximal landing zones in the aortic arch were classi ed as 0 to 4 according to Ishimaru's classi cation [10].

Statistical analysis
When comparing continuous variables, the Student's t-test and Wilcoxon test were used for parametric and nonparametric data, respectively, and are presented with the mean ± standard deviation (SD) or as median and interquartile range (IQR).Categorical variables were reported as absolute numbers or percentages and the Fisher's exact test or Chi-square test was used for comparison.The Kaplan-Meier method was used to estimate survival.For statistical analyses, p-values < 0.05 were deemed signi cant.Univariate and multivariate logistic regression models were utilized to determine independent risk factors.
To reduce the effect of selection bias and potential confounding in this retrospective cohort study, estimated propensity scores were used to match two groups.This was computed for each patient using a logistic regression model including the following variables: age, proximal maximal aortic size, aortic pathology, and proximal aortic tear site.The propensity score model was well-calibrated (Hosmer-Lemeshow goodness-of-t test; p = 0.784) with good discrimination (c-statistic = 0.712).To neutralize the effects of confounding variables, 35 patients in the open repair group were matched to 35 patients who underwent TEVAR using propensity score matching (PSM).The data were analyzed using SAS/STAT software, v. 9.4 (SAS Institute Inc., NC, USA) and SPSS 25 (IBM, Armonk, NY, USA).

Baseline characteristics of patients
Baseline characteristics of patients who underwent open repair and TEVAR are shown in Table 1.The mean follow-up duration was 70.2 ± 51.9 months (range, 0.0-212.0months).The median ages of the patients were 56.0 (range, 43.0-64.0)years, and 65.0 (range, 57.0-72.0)years for the open repair and TEVAR groups, respectively, which was signi cantly different between the two groups (p < 0.001).
Moreover, connective tissue diseases were observed at a signi cantly higher rate in the open repair group than in the TEVAR group (p < 0.001).After PSM, the baseline characteristics of the patients in each group exhibited no signi cant difference.TEVAR: thoracic endovascular aortic repair; SBP: systemic blood pressure; CAOD: coronary artery occlusive disease; PAOD: peripheral artery occlusive disease; COPD: chronic obstructive pulmonary disease.
Descending thoracic aortic diseases details of patients are shown in Table 2.No variables showed signi cant difference between the groups, before and after PSM, respectively.Operative details are showed in Table 3.In the open repair group, circulatory arrest perfusion was performed most frequently (60.0%), and thoracotomy was the most common approach (60.0%).In the TEVAR group, zone 3 TEVAR was performed most frequently (40.7%).Acute kidney injury (AKI) was signi cantly higher in the open repair group than those in the TEVAR group, before and after PSM, respectively.Bleeding, pulmonary and wound complications were signi cantly more observed in the open repair group; however, there was no statistical difference after PSM.There was no statistical difference in-hospital mortality and 30-day mortality between the two groups, before and after PSM, respectively.In-hospital mortality was observed in ve patients in the open repair group (11.1%), of whom four and one died of acute aorta-related complications and hospital-acquired pneumonia, respectively.In the TEVAR group, among three patients (3.3%), two patients died of delayed rupture after stent insertion and one patient died of aspiration pneumonia.

Change of aorta size after the repair
Figure 3 shows the changes in the aorta size after the repair.During the following 5 years, the maximal aortic diameter was reduced in the open repair group compared to that in the TEVAR group before and after PSM (p = 0.004 and p = 0.05).The maximal aortic diameter decreased from 55. 4  The cumulative survival of all-cause death was not statistically different, either before or after PSM, Additionally, the cumulative survival rates from aorta-related deaths were also not signi cantly different, before and after PSM, respectively (Fig. 4A,B).
A multivariable Cox proportional hazard model identi ed that age > 80 years, systolic blood pressure < 90 mmHg, diabetes, preoperative chronic renal failure, and aortic arch involvement were predictive factors in the in overall series; after PSM analysis, age > 80 years, and aortic arch involvement (p = 0.024 and p = 0.048, respectively) were independent predictors of aorta-related mortality (Table 7).

Discussion
Since, the report on the rst successful open repair of a descending thoracic aortic aneurysm with a prosthetic graft in 1953 by De Bakey and Cooley [11], an open surgical repair for treating descending thoracic aortic disease has been the gold standard for 50 years [11][12][13][14].Despite remarkably improved operative techniques and maximized organ protection, open repair of the descending thoracic aorta is still associated with high complications, including intraoperative and postoperative death, hemorrhage, stroke, and paraplegia [9,15].Dake et al. [12] proposed an alternate method of TEVAR which sought to provide better clinical outcomes in patients who were deemed to be at high risk for open repair or were typically considered nonsurgical candidates.Therefore, TEVAR has shown signi cantly improved early quality of life versus open repair and a general trend toward better short-term perioperative survival and freedom from major complications [1,3,4,16].However, TEVAR has anatomic restrictions such as severe thoracic aortic tortuosity, short landing and sealing zones, and extensive mural thrombus.These are the limiting factors, although a seemingly in nite variety of debranching and bypass procedures can be applied to extend either the proximal or distal sealing zones [6,17].Furthermore, signi cant complications related to stentgrafts were always implied [3,5,15].
Patients who underwent TEVAR have a tendency to have a worse prognosis and older age, with multiple comorbidities, than patients who underwent open repair.Due to the relative lack of data supporting the long-term reliability of TEVAR, open repair procedure has been preferentially offered to younger patients [4,18].Therefore, to neutralize the effects of age difference which could potentially unmask a mortality bene t, PSM was performed between the two groups to perform an adequately powered analysis.
In present study, operative time, postoperative length of stay, and procedure-related complications showed better results with TEVAR before and after PSM.Not surprisingly, TEVAR was considering as the procedure involved no aortic cross-clamping, ischemic time, or thoracotomy [4].In open repair cases, some disadvantages of deep hypothermia, including coagulopathy which caused di culty in controlling bleeding, retraction injury to a heparinized lung, cold injury to the lung, and a profound in ammatory response from the bypass circuit [19].For the in-hospital mortality of the open repair group, in present study, one patient died of pneumonia.AKI is another important complication and regarded as a marker of increased early or late morbidity and mortality after open repair or TEVAR [20].Patients who underwent TEVAR were older and tended to receive larger amounts of contrast agent, which was not safe considering the risk of AKI.In the present study, postoperative AKI was higher in the open repair group (42.2%), and dialysis was performed in 13.3% of patients.Eighteen patients who underwent TEVAR (19.8%) had an AKI with six requiring dialysis.Before and after PSM, AKI was higher in the open repair group; however, there was no statistical difference in dialysis between the two groups.
Although a short-term hospital outcome is more favorable for TEVAR, aorta-related complications are more frequent for TEVAR.Five patients (11.1%) in the open repair group underwent reintervention, and the most common cause of reintervention was new-onset aortic dissection or expansion.In the case of TEVAR, the most common cause of reintervention was endoleaks.Twenty-two (48.9%) patients who underwent reintervention showed no in-hospital mortality in the TEVAR; however, seven patients showed late mortality, one patient died of ABF and one patient died of sepsis due to stent-graft infection.
Ascending aortic replacement was performed in one patient with retrograde aortic dissection, four patients with ABF underwent aorta replacement and lobectomy, and one patient with AEF underwent aorta replacement and esophagotomy, and showed late mortality.Additionally, eight patients who were initially offered TEVAR, later crossed over to open repair due to di cult anatomy or other reasons.
Moreover, although long-term survival of the two groups had no signi cant difference, more reinterventions occurred in the TEVAR group; the costs of additional graft modules to treat endoleaks and of follow-up computed tomography increase hospital cost, attributing to the disadvantage of TEVAR.In addition, TEVAR has less procedure-related complications than that of open repair; patients had more adverse events, such as re-dissection, stula formation and stent-graft infection should be considered in the choice of approach.
Some authors have proposed that TEVAR does not change the natural history of the disease, and although less invasive, may be inferior to open therapies [24].In our present study, maximal aortic size decreased more in the open repair group than in the TEVAR group, but not dramatically.This supports the report that it does not alter natural history of aortic pathologies, and, emphasizes the importance of longterm follow up.For this reason, in patients requiring TEVAR, the establishment of a precise TEVAR indication will reduce the requirement for further reintervention; better results are expected with improvements in debranching skills and stent-graft development.
Our study has several limitations.First, it was a single-center retrospective study that included a small number of patients, with a possible selection bias that might have affected our results.Second, the difference of follow-up duration and frequency can affect the survival rate of both groups.We performed PSM attempts to reduce the bias due to confounding variables.However, since TEVAR was introduced in 2007, it has a relatively short follow-up duration, whereas more frequent follow-up to monitor stent-grafts is expected to affect the results.Finally, the functional status of patient information in uenced treatment strategy; there was no data interpretation, and studies on cost analysis, which is an increasingly important consideration for treatment strategy, have not been conducted.

Conclusions
In conclusion, our study showed that a long-term comparison between the open repair and TEVAR group demonstrated similar results in patients with descending aortic syndromes.However, patients who underwent TEVAR showed superior short-term recovery results and a higher reintervention rate than the open repair group.Larger multicenter population studies that consider quality of life could support our present study results.

Declarations
Ethics approval and consent to The institutional review board (IRB) of Kyungpook National University Hospital approved this retrospective study and waived the requirement for individual patient consent (IRB approval No. 2019-10-051).Freedom from aortic reintervention before propensity matching and after propensity matching  Cumulative survival of all-cause death and aorta-related death

Figures Figure 1 Page 23
Figures

Figure 3 Changes
Figure 3

Table 3
Postoperative outcomes and complications are shown in Table4.The patients in the open repair group signi cantly required more operative time, needed longer ventilator care, stayed longer in the intensive care unit, and had longer periods of hospitalization than those in the TEVAR group (all p < 0.001, respectively), before and after PSM, respectively.
Aorta-unrelated death secondary to cancer occurred in three patients and three patients died of pneumonia.Late deaths secondary to unknown causes occurred in four patients.In the TEVAR group, overall mortality was 35.2% (32/91), and aorta-related mortality was 12.1% (11/91) during follow-up.Aorta-unrelated death secondary to cancer occurred in three patients and two patients died of pneumonia.Late deaths secondary to unknown causes occurred in 10 patients.

Table 7
Cox proportional hazard regression analysis for aorta-related mortality