Management of Anomalous Left Vertebral Artery During Open Total Arch Replacement

Abstract

Background

An anomalous left vertebral artery (aLVA) can complicate aortic arch surgery. We examined the safety of various aLVA revascularization strategies during open total arch replacement.

Methods

We retrospectively evaluated 92 patients undergoing total arch replacement from January 2018 to May 2023 and identified 11 patients with aLVA. A comparison group (n = 31) with conventional 3-branched anatomy was selected by 3:1 matching on age, sex, circulatory arrest time, and operative mortality. Forty-two patients were selected for analysis. The primary outcome was perioperative stroke within 30 days. Secondary outcomes included spinal cord ischemia and long-term stroke.

Results

Patients with aLVA had an average age of 53 ±16 years. Indications included Stanford type A dissection (n = 4), chronic dissection with aneurysmal degeneration (n = 4), and primary aneurysmal disease (n = 3). The aLVA was reconstructed by transposition to the left carotid (n = 2) or subclavian (n = 8). In 1 case, partial zone 2 arch replacement was performed proximal to the aLVA and manipulation was not required. Total arch replacement with frozen elephant trunk was performed in 8 of 11 cases. No postoperative mortality was observed. One patient experienced transient postoperative stroke (9% vs 9%, P = 1.00). One patient received lumbar drain for suspected spinal cord ischemia (9% vs 7%, P = 0.59). One patient experienced stroke 6 months post operation. One-year patency of transposition was 100%.

Conclusions

The presence of aLVA does not impact outcomes of open arch surgery. Our strategy to preserve the aLVA may be preferable to simple ligation. We describe several safe and feasible approaches for reimplantation of this anomalous variant.

In Short.

• ▪We describe several approaches for reimplantation of an anomalous vertebral artery with excellent patency and low stroke rates.
• ▪Preservation of anomalous left vertebral artery is safe and feasible during total arch replacement.

Anomalous origin of the left vertebral artery (aLVA) is the second most common arch variant and occurs in 2%-8% of the population.¹ The left vertebral artery typically originates as the first branch of the left subclavian artery (Figure 1), however, it can aberrantly arise from the aortic arch or carotid artery. Origin from the arch poses specific risks for aortic arch operations because an aLVA is associated with increased risk of stroke when not properly revascularized.²

Figure 1.

Anomalous origin of left vertebral artery. The left vertebral artery arises directly off of the aorta, either proximal (pictured) or distal (unpictured) to the left subclavian artery.

The optimal technique to manage aLVA during open aortic arch surgery is not described in the literature, although it is hypothesized that revascularization may reduce the risk of stroke and spinal cord ischemia. The Society of Vascular Surgery gives a class I indication to revascularize the left subclavian, from which it can be extrapolated that the left vertebral artery should be revascularized as well.^3,4 We describe several techniques to manage an aLVA. The objective of this study was to evaluate the safety and stroke risk of these strategies.

Patients and Methods

We retrospectively examined patients undergoing total aortic arch replacement (TAR) from Jan-2018 to May 2023. Patients with aLVA were compared to a control group with conventional arch anatomy. This study was approved by our institutional review board including consent for publication (IRB STU00012288).

The Society of Thoracic Surgeons database definitions were used for variables when available. Additional record review was done to obtain long-term stroke outcomes, intraoperative events, and follow-up. The primary outcome was perioperative stroke within 30 days. Secondary outcomes included spinal cord ischemia and long-term stroke rates. Postoperative patency of the left vertebral artery was verified by computed tomography angiogram.

All patients in this series had an intraoperatively verified aLVA. TAR was performed via median sternotomy under total cardiopulmonary bypass using either axillary or central cannulation. Proximal aortic reconstruction was performed during cooling to deep hypothermia. Cerebral protection used antegrade cerebral perfusion except for 1 patient in the aLVA group and 3 patients in the control group, for whom we used retrograde cerebral perfusion. The control group had a similar distribution of cerebral protection strategies. TAR was performed with a zone 2 anastomosis and frozen elephant trunk stent graft or branched Dacron graft (DuPont) alone. The configuration of the aLVA revascularization was guided by a goal to avoid disruption of the left vertebral artery and maximize anatomic fit. Postoperatively, patients receive aspirin for thrombosis prevention.

We applied descriptive and comparative statistics to analyze the data. Comparison of categorical variables used χ² or Fisher’s exact test, whenever cell counts were less than 5. Two-sided P values <.05 were considered significant. Patient matching used the MatchIt package in R (R Foundation for Statistical Computing). Matching variables (age, sex, operative mortality, and circulatory arrest time) were selected based on content knowledge and designed to be used for simple risk assessment. All statistical analyses used R 4.3.1.

Results

In total, 11 of 92 (12%) patients were identified to have an aLVA during TAR. After 3:1 matching, 31 patients were selected as controls (Table 1). Most patients underwent redo sternotomy (60%). Operative indications between aLVA and conventional arch group were not significantly different (P = .44). Anomalous anatomy was identified preoperatively in 5 of 11 patients.

Table 1.

Preoperative Characteristics

Characteristic	Anomalous Left Vertebral Artery (n = 11)		Normal Arch (n = 31)		P Value
Age, y	53.2	±16.3	60.3	±12.4	.21
Body mass index	32.5	±6.8	27.3	±5.0	.02
Male	11	(100)	22	(71)	.11
Diabetes	2	(18)	1	(3)	.33
Creatinine	1.04	(0.96, 1.15)	1.1	(0.86, 1.25)	.95
Lung disease	0	(0)	4	(13)	.51
Peripheral vascular disease	2	(18)	4	(13)	.99
Prior stroke	2	(18)	6	(19)	>.99
Ejection fraction	59	(54, 65)	59	(55, 66)	.98
NYHA class III/IV	0	(0)	1	(2)	>.99
Coronary artery disease	0	(0)	7	(22)	.21
Atrial fibrillation	1	(9)	6	(19)	.75
Prior sternotomy	6	(54)	19	(61)	.97
Connective tissue disease	4	(36)	2	(7)	.58
Indication
Acute dissection	4	(36)	11	(36)	.44
Chronic dissection	4	(36)	6	(19)
Aneurysm	3	(27)	14	(45)

Values are presented as mean ± SD, n (%), or median (interquartile range).

NYHA, New York Heart Association.

Reconstruction of the aLVA occurred as follows (Figure 2): (1) Reattachment to left carotid (n = 2). (2) Transposition to native left subclavian (n = 4), subclavian graft (n = 3), or as en bloc button (n = 1). (3) A partial zone 2 arch replacement was performed proximal to the aLVA (n = 1), thereby avoiding manipulation of the vessel. TAR was performed with frozen elephant trunk in 8 of 11 cases. In the conventional arch group, frozen elephant trunk was performed in 23 of 31 cases (74.2% vs 74.1%).

Figure 2.

Typical repair configuration of anomalous left vertebral artery. (A) Transposition to the left subclavian graft. (B) Transposition to native subclavian artery. (C) Reattachment of left subclavian artery and anomalous left vertebral artery as a button. (D) Transposition to left carotid artery. (E) Modified zone 2 aortic anastomosis proximal to anomalous branch origin.

Intraoperative variables are shown in Table 2. The median bypass time was 241 minutes (interquartile range [IQR], 210, 273), higher for the aLVA group (282 [243, 311] vs 229 [204, 250]). The median cross clamp time was also higher in the aLVA group (225 [155, 254] vs. 171 [139, 225]). Average circulatory arrest time was 53 minutes. One patient in the aLVA group experienced renal failure requiring temporary dialysis.

Table 2.

Intraoperative Details and Operative Outcomes

Operative Details	Anomalous Left Vertebral Artery (n = 11)		Normal Arch (n = 31)		P Value
Bypass time, min	282	(243, 311)	229	(204, 250)	.01
Cross-clamp time, min	225	(155, 254)	162	(140, 215)	.14
Circulatory arrest time, min	53	(47, 55)	52	(37, 65)	.72
Cerebral perfusion strategy					.37
Retrograde	1	(9)	3	(10)
Unilateral antegrade	4	(36)	5	(16)
Bilateral antegrade	6	(54)	23	(74)
Frozen elephant trunk	8	(73)	23	(73)	>.99
Follow up time, d	728	±613	823	±601	.58
Spinal cord ischemia	1	(9)	2	(7)	.59
Stroke
Perioperative stroke	1	(9)	3	(9)	>.99
Permanent stroke	1	(9)	1	(3)	.83
Late stroke (>30 d)	1	(9)	1	(3)	.99
Patent vertebral at follow-up	11	(100)	…	…	…
Long-term all-cause mortality	1	(9)	1	(3)	.99

Values are presented as n (%), mean ± SD, or median (interquartile range) unless otherwise specified.

Average length of follow up was 2.0 ± 1.7 years. In the aLVA group, a perioperative infratentorial stroke occurred in 1 patient with temporary expressive aphasia. Suspected spinal cord ischemia with lower extremity weakness occurred in 1 case, which improved with lumbar drain. No cases required reoperation. Late stroke occurred in 1 patient at 6 months. The patient presented with a facial droop and was diagnosed with a transient ischemic attack affecting the posterior region. The 1-year patency of transpositions was 100%. In the control group, perioperative stroke was observed in 3 patients, of which 1 developed motor deficit. The transient strokes were determined to be related to central embolic phenomena due to arch manipulation, and the remaining stroke was an ischemic posterior stroke. Spinal cord ischemia occurred in 3 cases, resulting in 2 additional patients with motor deficit. One patient experienced a transient cranial nerve injury of the vagus nerve affecting swallowing.

Comment

In this study examining patients with an aLVA during open TAR, we explore several methods to address this challenging anatomy. Despite an increase in bypass and cross-clamp times, we observed no operative mortality or morbidity. Thus, reattaching an aLVA to the nearby left carotid or subclavian are both feasible options, especially when attached to the branch of a frozen elephant trunk graft, which would eliminate the risk of left subclavian coverage in future descending aortic intervention.⁵

There is no specific guidance regarding whether an aLVA should be revascularized during open arch procedures. In our cohort, aLVA was present in 12% of patients undergoing TAR. The risk of perioperative stroke or spinal cord ischemia when ligating an aLVA is estimated between 6% and 10% based on studies examining left subclavian artery coverage during thoracic endovascular aortic repair.⁶ It is unknown how often surgeons elect to ligate or revascularize an aLVA. A few series have examined revascularization by a staged transposition or bypass during endovascular or hybrid arch procedures with planned subclavian coverage observing low stroke rates and excellent vessel patency at 1 year.⁷ In our center, we routinely revascularize an aLVA during TAR. Although 1 patient experienced transient ischemic attack affecting the posterior region, follow-up imaging demonstrated a chronic right vertebral occlusion with distal reconstitution and a slight narrowing of the reconstructed left vertebral artery. Thus, it is possible that the posterior symptoms were related to increased demand in the context of reduced supply in both arteries.

Revascularization adds additional technical challenges and increased circulatory arrest time. For example, the subclavian is challenging to access from a median sternotomy as it is deep and recessed. Furthermore, the caliber of the vertebral artery is small (3-4 mm), which is difficult to manipulate and reimplant. There have been few case series examining revascularization of the left vertebral artery when it arises directly from the aorta. The largest series in patients undergoing total arch replacement examined 21 patients with an aLVA who presented with type A dissection.⁸ They described using a frozen elephant trunk technique with individual branch grafts, similar to our series. On the other hand, a few case series in the neurosurgical literature describe that vertebral artery ligation may be safe in patients with complete posterior circulation and absence of additional occlusive disease. Since manipulation of the aortic arch may cause embolic phenomena, it is unknown if this approach would be beneficial in arch cases.⁹ In our series, we present several configurations and techniques that can also achieve procedural success. Thus, our findings support that a single stage approach is safe, practical, and feasible during aortic arch surgery. The benefits of a single stage approach include a reduced need for dedicated preoperative imaging of the cervical anatomy as has been suggested, and also does not accumulate the risks of multiple operations. Future guidelines should consider discussing the impact of an aberrant left vertebral artery on aortic arch surgery.

There are other methods to address anomalous left vertebral anatomy. If identified preoperatively, some elect for 2-stage revascularization, especially because a supraclavicular incision provides exposure and minimizes technical challenges. The disadvantage of this approach occurs because additional operative planning is required and, with emergency procedures such as dissection, this could not feasibly be performed, as was the case with 40% of cases in this series. Furthermore, anomalous anatomy may only be detected at time of operation due to suboptimal preoperative imaging or scans, which would limit the possibility for a 2-staged approach. Finally, for 1 patient in our cohort, a zone 2 anastomosis was made proximally to the aLVA, and manipulation was avoided. However, it is important to recognize that this decision has implications for subsequent endovascular intervention. Additional revascularization techniques also include a subclavian vertebral artery bypass with a saphenous vein graft, which would limit the amount of mobilization required to revascularize the vertebral artery. Finally, a few have described a left internal mammary to aLVA anastomosis. The disadvantage is that this conduit is not always present and is challenging to use in acute settings.

Limitations

This is a retrospective study with a small number of cases, which is inherent in a series examining rare anatomic variants. In addition, there may have been minor differences in the operative technique of the surgeons who performed the cases. The strengths of this series include excellent clinical follow-up through our comprehensive aortic center and consistent operative approach to management of this aberrant anatomy.¹⁰

Conclusions

Preservation of aLVA is feasible during open TAR. These arteries have excellent patency. Stroke rates were low in this series without added morbidity. We describe several safe approaches for reimplantation of the anomalous vertebral artery. Future studies may determine if there is a protective effect of aLVA transposition at time of operation.