Surgery for chronic type B dissection with aneurysmal degeneration

Abstract

Purpose

Open repair of descending thoracic or thoracoabdominal aortic aneurysm (TAAA) continues to carry a not insignificant operative risk, even in experienced hands. Over the past three decades, there has been considerable improvement in both the mortality and morbidity associated with these procedures. Herein, we describe our operative results and long-term outcomes in patients with chronic type B aortic dissections.

Methods

Review of the aortic surgical database was conducted to identify all consecutive patients who underwent repair of TAAA for chronic type B dissection from May 1997 to March 2018. The primary end point was operative mortality with secondary end points as the composite of major adverse events as well as each of the individual complications.

Results

One hundred and fifty-three patients met inclusion criteria with 54.9% (84/153) having surgery on an elective basis. The mean age was 58.9 years with a majority of male gender—107/153 (69.9%). Eighty-three (54.2%) of the TAAA were extent I, while 36 (23.5%) were extent II and 34 (22.3%) extent III-IV. Operative mortality was 8.5% (13/153) with eight of the deaths in patients who presented with extent II TAAA. On Kaplan-Meier survival analysis, 87.5% (95% confidence interval (CI) 77.9–97.1%) of the elective cohort were alive after 5 years while only 69.9% (CI 55.2–84.6%) of those in need of urgent/emergency intervention survived (p = .039).

Conclusions

In a majority of patients with chronic type B dissections, reproducibly, excellent outcomes can be achieved with relatively low risk of mortality. In the higher risk subsets of patients with extent II TAAA, careful consideration and discussion of expected outcomes will help inform the decision-making process.

Introduction

The treatment paradigm for acute type B aortic dissections continues to evolve. While many patients with type B dissections, especially those with malperfusion syndromes, are treated with endovascular stent-grafting, a large proportion of patients presenting continue to be managed medically [1, 2]. In a subset of patients, surgery is emergently necessary to treat an acute aortic catastrophe, such as free rupture or malperfusion. Those who are treated endovascularly or managed medically are at risk for aneurysmal degeneration of the dissected aorta, which may require future repair. Open repair of descending thoracic or thoracoabdominal aorta continues to carry a not insignificant operative risk, even in experienced hands. However, given the natural propensity of compromised aortic tissue towards continued dilatation and eventual rupture, the chance of surgical success in those with expanding aneurysms outweighs the risk of non-operative management [3, 4]. The morbidity and mortality associated with open surgical repair are far from trivial, and the presence of multiple false lumens increases technical complexity [5]. The operative mortality risk remains 4 to 10% even in the most experienced hands [6–8].

Over the past three decades, there has been considerable improvement in both the mortality and the morbidity associated with open descending and thoracoabdominal aortic procedures. The use of circulatory support, end-organ preserving techniques, and spinal protection techniques in nearly all patients along with a dedicated multi-disciplinary team continues to portend better results [9, 10]. Herein, we describe our operative results and long-term outcomes in patients with chronic type B aortic dissections.

Materials and methods

Patient population, definitions, and end points

This study was approved by the institutional review board. The need for individual patient consent was waived. Review of prospectively collected data from the Weill Cornell Medicine Department of Cardiothoracic Surgery aortic surgery database was conducted to identify all consecutive patients who underwent repair of TAAA for type B chronic dissection from May 1997 to March 2018. The database is constantly updated and maintained by a team of clinical information analysts; data collection is validated by means of external and internal control. Preoperative and perioperative variables are entered prospectively during in-hospital stay. Postoperatively, clinical and computed tomographic evaluations are performed every year and in case of clinical symptoms suggestive of disease progression, and data are entered at the time of the follow-up visit. In case of missing or unreliable data, direct interview with the patient, a relative, or the treating physician is performed.

The primary end point was operative mortality, defined as death within the same hospitalization of the index surgical procedure or within 30 days postoperatively. Secondary end points were the incidence of major postoperative complications, including reoperation for bleeding, postoperative myocardial infarction, postoperative stroke, need for tracheostomy, spinal cord injury (paralysis and paraparesis), new onset renal insufficiency requiring renal replacement therapy, left recurrent laryngeal nerve injury requiring intervention, and a composite of major adverse events (MAE). MAE was composed of operative mortality and the five major postoperative complications.

Aortic dissections were classified according to DeBakey typology [11] and TAAAs according to the Crawford classification system [3]. Our criteria for surgical treatment of type B dissections are as follows: aneurysmal degeneration with a size threshold cutoff of 5.5 cm or growth more than 4 mm per year, rupture, shock, neurologic compromise, refractory pain, refractory hypertension, end organ, or lower extremity malperfusion.

Surgical technique

The details of our surgical procedure have been previously published [12]. Briefly, a fifth, sixth, or seventh intercostal space thoracotomy or thoracoabdominal incision was used. Resection of adjacent ribs was performed as needed for additional surgical exposure. Based on the extent and complexity of the planned aortic reconstruction, either systemic left heart bypass (LHB) or organ-specific (warm hematinic mesenteric and cold crystalloid renal) perfusion adjuncts were used, according to existing evidence [13, 14]. The primary perfusion technique for type B dissections with multiple patent intercostal arteries or complex luminal configurations was left heart bypass. With more limited disease, if perfusion adjuncts were deemed unnecessary, a clamp-and-sew technique was employed. In general, our approach to organ protection was the following: extent I TAAA with selective LHB, extent II with TAAA with customary LHB plus mesenteric and renal perfusion, extent III TAAA with customary mesenteric and renal perfusion with or without LHB, and extent IV TAAA with selective mesenteric perfusion plus customary renal perfusion. In patients with compromised left ventricular function, LHB was also utilized to offload the left ventricle and facilitate aortic clamping. For extent I and II TAAAs, the site of proximal clamping was preferentially distal to the takeoff of the left subclavian artery. In a minority of cases, proximal clamping distal to the takeoff of the left common carotid artery was not an option and the use of cardiopulmonary bypass and deep hypothermic circulatory arrest was necessary.

The extent of resection depended on aortic size, symptoms/complications, and sporadic versus familial pathologic process. Non-aneurysmal segments were not replaced but were fenestrated to ensure adequate perfusion of both lumens. This allowed us to preserve the maximal amount of segmental arteries and thus minimize spinal cord ischemia.

Reinfusion of shed blood was performed with a warm rapid infusion system (Belmont Instrument Corporation, Billerica, MA). When LHB or the clamp-and-sew technique was used, core temperature was allowed to decrease to 33 °C before cross-clamping the aorta. Re-implantation of intercostal arteries was performed with the inlay-inclusion technique. Visceral and renal arteries were either re-implanted or bypassed as dictated by anatomy. Hemashield Dacron grafts (Macquet Corporation, Oakland, NJ) were used. Preoperative spinal drain insertion was attempted in all hemodynamically stable patients. In unstable patients, spinal drain insertion was performed immediately after surgery before transporting the patient to the intensive care unit. Drains were maintained with an intrathecal pressure less than 12 cm H₂O for 72 h, and mean arterial pressure was maintained at greater than 85 mmHg.

Statistical analysis

Data were stored using Microsoft Access 2010 software (Microsoft, Redmond, WA) and analyzed using IBM SPSS Statistics version 22 (IBM, Armonk, NY) and R version 2.15.2 (R Foundation for Statistical Computing). Backward stepwise logistic regression analysis was attempted to identify specific preoperative predictors, but there were too few data points to provide meaningful results. Postoperative survival was assessed with the Kaplan-Meier method.

Results

During the study period, a total of 574 patients underwent open TAAA repair at our institution. From this population, we identified 153 patients who underwent operation due to aneurysmal degeneration from chronic type B aortic dissection.

Preoperative characteristics

Preoperative variables are described in (Table 1). The mean age of the study cohort was 58.9 ± 14.4 years with 69.9% males (107/153) while the mean ejection fraction was 49.5 ± 7.7%. The majority of the patients had a history of hypertension (94.8%) and smoking (68.0%). A minority of patients presented with chronic obstructive pulmonary disease (28.1%), connective tissue disorders (22.9%), peripheral vascular disease (13.1%), previous coronary revascularization (11.1%), previous cerebro vascular accident (10.5%), and diabetes (3.9%).

Table 1.

Preoperative

Variable	Chronic type III (n = 153)
Age (mean, std. dev.)	58.9 ± 14.4
Males	107 (69.9)
Smoking	104 (68.0)
Previous coronary revascularization	17 (11.1)
Hypertension	145 (94.8)
COPD	43 (28.1)
CVA	16 (10.5)
Connective tissue disorders	35 (22.9)
PVD	20 (13.1)
Diabetes	6 (3.9)
Procedure status
Elective	84 (54.9)
Urgent/emergent	69 (45.1)
Paraparesis/paraplegia	4 (2.6)
Previous OHS	67 (43.8)
Arterial occlusive disease	12 (7.8)
Concomitant procedures	43 (28.1)
Aneurysm size (cm; mean, std.dev.)	6.9 ± 1.4
Aneurysm extent (TAAA)
I	83 (54.2)
II	36 (23.5)
III-IV	34 (22.3)
Ejection fraction (mean, std. dev.)	49.5 ± 7.7

Data presented as n (%), unless otherwise noted

COPD chronic obstructive pulmonary disease, CVA cerebrovascular accident, PVD peripheral vascular disease, OHS open heart surgery, TAAA thoracoabdominal aortic aneurysm

Of the 153 patients, 84 (54.9%) had an elective procedure while 69 (45.1%) were repaired on an urgent or emergent basis. Sixty-seven patients (43.8%) had previous open heart surgery while only 12 (7.8%) had arterial occlusive disease and 4 (2.6%) had preoperative paraparesis or paraplegia. Forty-three of the 153 patients (28.1%) underwent concomitant procedures.

Intraoperative details

The majority of the descending dissections were extent I (54.2%), while there were 36 (23.5%) extent II and 34 (22.3%) extent III-IV aneurysms (Table 1). As seen in (Table 2), cerebrospinal fluid drainage was successfully achieved in 139 of 153 patients (90.8%). At least one set of intercostal arteries was re-implanted in 103 patients (67.3%) while renal and/or visceral perfusion was employed in 46 patients (30.1%). Left heart bypass was used in approximately half of all patients, 77/153 (50.3%), and deep hypothermic circulatory arrest with cardiopulmonary bypass was needed in 7 patients (4.6%).

Table 2.

Intraoperative

Variable	Chronic type III (n = 153)
CSF drainage	139 (90.8)
Intercostal reimplantation	103 (67.3)
Renal/visceral perfusion	46 (30.1)
Partial bypass	77 (50.3)
Circulatory arrest	7 (4.6)

Data presented as n (%), unless otherwise noted

CSF cerebrospinal fluid

Postoperative outcomes

As enumerated in (Table 3), operative mortality was 8.5% (13/153). A majority of these occurred in patients with extent II aneurysms, and in this group, the mortality was 22.2% (8/36). Three patients required reoperation for bleeding (2.0%) while one patient each suffered a stroke (0.7%) and myocardial infarction (0.7%). Ten patients required postoperative tracheostomy (6.5%) while nine patients suffered from new-onset renal insufficiency requiring renal replacement therapy (5.9%). Twelve patients needed intervention following left recurrent nerve injury (7.8%). Four patients had paraparesis postoperatively (2.6%) with one patient suffering paraplegia (0.7%).

Table 3.

Postoperative results

Variable	Chronic type III (n = 153)
Operative death	13 (8.5)
MAE	28 (18.3)
Take back for bleeding	3 (2.0)
Myocardial infarction	1 (0.7)
Periop CVA	1 (0.7)
Tracheostomy	10 (6.5)
Postop spinal cord status
Paraparesis	4 (2.6)
Paraplegia	1 (0.7)
Dialysis/CVVH/UF	9 (5.9)
Left recurrent nerve repair required	12 (7.8)

Data presented as n (%), unless otherwise noted

MAE major adverse events, CVA cerebrovascular accident, CVVH continuous veno-venous hemofiltration, UF ultrafiltration

On Kaplan-Meier survival analysis, patients undergoing elective surgery as compared with those needing urgent/emergency procedures had a statistically significant survival advantage. At 5-years postoperatively, 87.5% (95% confidence interval (CI) 77.9–97.1%) of the elective cohort were alive while only 69.9% (CI 55.2–84.6%) of those in need of urgent/emergency intervention survived (p = .039) (Fig. 1).

Fig. 1.

Survival

Discussion

Repair of chronic dissection of the descending aorta remains a complex and technically demanding operation. The surgical maneuvers necessary to complete the repair in part depend on the etiology of the aneurysm formation, whether it is due to degenerative/atherosclerotic disease, chronic type I dissection, or chronic type III dissection [8]. In this report, we focused on dissections which originated distal to the left subclavian artery and propagated antegradely. We sought to evaluate the characteristics of this patient population and report on our results from a high volume aortic center.

Unlike degenerative aneurysms, which are seen in a slightly older population, TAAAs due to chronic type B dissection are seen in a younger group and the mean age in this series was 58.9 years. A significant proportion (22.9%) of these patients has a connective tissue disorder, which predisposes them to getting the initial aortic dissection and subsequent aneurysmal expansion. The area of the aorta at highest risk for aneurysmal growth is the proximal descending thoracic aorta, distal to the origin of the left subclavian artery. This is most often where the primary tear originates, and this proximal segment of the aorta experiences higher forces than the downstream aorta. As a result, 54.2% of the group had an extent I TAAA repair and extent II-IV were less common.

Over half of the patients were repaired with circulatory support. Unlike patients with atherosclerotic aneurysms, who often have many occluded intercostal arteries due to atheromatous burden or mural thrombus, patients with chronic aortic dissections often have many large, patent, heavily backbleeding intercostal arteries. Additionally, the anatomy can be difficult to discern due to multiple lumens and fenestrations. The oversewing of the many intercostal arteries and establishment of the proper anatomy takes additional operative time, and the use of partial bypass allows for perfusion of the distal aorta and safe extension of cross-clamp times, while the proximal aortic reconstruction is being done. Our low rates of postoperative end-organ dysfunction and spinal cord injury help to justify this approach. Very few patients were repaired under circulatory arrest as we reserve this technique for only those in whom we cannot place a cross-clamp either distal to the left subclavian artery or between the left internal carotid and subclavian arteries.

The operative mortality in this group with chronic type III dissections was 8.5%, which is higher than in other subgroups of patients undergoing TAAA repair (such as those for acute complicated type B dissection or those with connective tissue disease). A majority of these deaths (8/13) occurred in patients with extensive aneurysmal disease (extent II) requiring replacement of the entire distal aorta from the left subclavian artery to the aortic bifurcation. Extent II is the most technically demanding repair, and it requires complex reconstruction with re-implantation of multiple aortic branch vessels. More recently, we have transitioned to using a graft with four prefabricated branches for re-implantation of visceral vessels, and this has resulted in improved outcomes with complex repairs. The branched graft has allowed us to modify the order in which the repair is completed in order to restore blood flow to the spinal cord and distal vasculature more quickly.

Coselli et al. recently reported on their experience with chronic type B patients with Marfan syndrome [15]. While they reported an operative mortality in this cohort of only 4% (2/54), other major adverse events were similar to those observed in our series. Though they reported no perioperative strokes, they had a 6% rate of spinal cord deficits, an 11% need for tracheostomy, and a 4% incidence of reoperation for bleeding. Of note, 19 of the 54 patients (35%) suffered left vocal cord paralysis. The mean age of repair in this series was 41.1 ± 14.0 years of age, whereas our patients were in their sixth decade (58.9 ± 14.4 years old). Due to the earlier intervention necessitated by the natural history of Marfan syndrome, it is plausible that the younger age of those patients explains the lower mortality rate. Furthermore, more than 70% of the operations in the Houston cohort were performed in an elective setting while more than 45% of cases in our series were conducted urgently or emergently.

An additional factor that affects patient survival is hospital volume of cases. In a 2003 study of the National Inpatient Sample, the overall mortality of intact TAAA repair was 22.3%, with a significant difference based on both hospital volume and surgeon volume. The mortality in a high-volume hospital (defined as a median of 12 cases per year) was significantly lower than that of both medium-volume and low-volume hospitals (15% vs. 23.8 and 27.3%, p < .001 and p = .001, respectively) [16]. At our high volume center, our mortality rate of 8.5% is well below the reported national averages, even amongst the high volume cohort.

Repair of aneurysms associated with chronic type B aortic dissections continues to be a challenging surgical undertaking associated with a significant risk of morbidity and mortality. Increasing complexity of the disease process in terms of aneurysm extent and anatomic complexity leads to increased operative risk. Proper risk stratification is necessary to properly inform patients of their prognosis with surgery. In a majority of patients, reproducibly excellent outcomes can be achieved with relatively low risk of mortality. In the higher risk subsets of patients with extent II TAAA, careful consideration and discussion of expected outcomes will help inform the decision-making process.

Compliance with ethical standards

Conflict of interest

Dr. Leonard, Dr. Lau, Dr. Iannacone, Dr. Gaudino, Ms. Munjal and Dr. Girardi declare that they have no conflict of interest.

This study was approved by the institutional review board. The need for individual patient consent was waived.