This retrospective, single-center study included patients with aortic arch pathologies who underwent TEVAR at our institution between September 2010 and August 2023. Inclusion criteria were: (1) diagnosis of aortic arch pathology, including but not limited to acute aortic syndrome (AAS), aortic aneurysm, or related conditions; (2) planned Zone 2 TEVAR with coverage of the LSA, potentially involving coverage of the ILVA; and (3) preoperative computed tomography angiography (CTA) confirming the presence of an ILVA. Exclusion criteria were: (1) poor-quality CTA imaging that could not be reliably evaluated, and (2) patients deemed unsuitable for TEVAR due to other medical reasons. AAS refers to a spectrum of life-threatening conditions, including aortic dissection, intramural hematoma, and penetrating aortic ulcers, which are recognized indications for TEVAR under current clinical guidelines [2, 5]. This study was approved by the Ethics Committee of Fuwai Hospital. Owing to the retrospective design, the requirement for informed consent was waived.

Two experienced radiologists independently measured vascular diameters using Endosize software (Therenva, Rennes, France). To minimize measurement errors, diameters were assessed 2 cm from the origin of both vertebral arteries. Postoperative occlusion of the isolated left vertebral artery was defined as the absence of intraluminal blood flow following TEVAR. Postoperative stenosis of the isolated left vertebral artery was defined as a $\ge$30% reduction in diameter compared with preoperative measurements. Primary technical success was defined as successful deployment of the stent graft, excluding aortic pathology, with no conversion to open surgery and no surgery-related mortality. In the Fenestration Group, preservation of ILVA patency was additionally required to qualify as a technical success, reflecting the goal of revascularization.

All procedures were performed under general anesthesia. Standard disinfection and draping were applied to the bilateral femoral and left brachial artery regions. A 5F gold marker pigtail catheter was introduced through the femoral artery into the ascending aorta for digital subtraction angiography (DSA) (Fig. 1A). Measurements of the aortic lesion and its morphological characteristics were obtained from CTA and DSA images. The aortic stent graft was selected with a diameter oversized by 15–20% relative to the measured aortic diameter. Systemic heparinization was administered in all patients. An Ankura thoracic aortic-covered stent (Lifetech Scientific Co., Ltd., Shenzhen, Guangdong, China) was used in every procedure. This device features a longitudinally distributed metallic support structure that provides additional reinforcement along the greater curvature of the aorta and incorporates a radiopaque marker to delineate the transition between the bare metal and covered segments. When aligned perpendicularly to the X-ray beam, the marker appears as a “$\mathrm{\infty }$” symbol; and when parallel, it appears as a “—” symbol.

Fig. 1.

Procedural steps for the PMF technique in TEVAR. (A) Preoperative DSA showing aortic pathology and vertebral artery anatomy. (B) Fenestration of the ILVA and LSA openings performed with electrocautery, guided by the “$\mathrm{\infty }$” radiopaque marker for accurate positioning. (C) Deployment of the stent graft with subsequent confirmation of vessel patency on DSA. PMF, physician-modified fenestration; DSA, digital subtraction angiography; TEVAR, thoracic endovascular aortic repair; ILVA, isolated left vertebral artery; LSA, left subclavian artery.

Upon confirmation of the “$\mathrm{\infty }$” radiopaque marker, the outer sheath of the aortic stent was retracted 4–5 cm to expose its proximal end, and fenestration was performed using an electrocautery pen (Fig. 1B). The fenestration covered the openings of the left subclavian artery and isolated the left vertebral artery. The sheath was then advanced using the bunding technique until it was repositioned in its original location. The stent graft was subsequently delivered to the aortic lesion over a Lunderquist super-stiff guidewire (William Cook Europe, Bjaeverskov, Denmark). Deployment was performed with the covered segment positioned adjacent to the posterior edge of the left common carotid artery (Fig. 1C). A pigtail angiographic catheter was reintroduced to perform DSA and confirm patency of the supr-aortic branches. In the Coverage Group, the procedural steps were similar except that no fenestration or bridging stents were performed. Both the ILVA and LSA were intentionally covered during Zone 2 TEVAR without revascularization. Endoleaks detected intraoperatively or during follow-up were classified as follows: Type Ia, proximal endoleak caused by inadequate sealing at the landing zone; Type II, retrograde flow from branch arteries arising from the excluded segment, most commonly via the left subclavian artery; and Type III, leakage due to stent graft failure, such as component separation or fabric disruption [6]. To preserve LSA patency, we generally consider implanting a bridging stent (Fig. 2). However, in cases of marked tortuosity of the proximal subclavian artery, dissection involving the subclavian artery, or aneurysmal disease at its origin, bridging stent implantation was deemed unsuitable.

Fig. 2.

Illustration of TEVAR procedure with PMF and bridging stent. The diagram demonstrates the key steps of TEVAR with PMF for ILVA preservation and bridging stent placement for the LSA.

All patients with a history of hyperlipidemia or diabetes mellitus received standard lipid-lowering and antidiabetic treatment during the perioperative period, applied consistently across both groups. In the Fenestration Group, antithrombotic therapy was tailored according to the diameter of the bridging stent: patients with stents $\le$8 mm received dual antiplatelet therapy (DAPT) for 3 months postoperatively, while those with stents $>$8 mm received single antiplatelet therapy. Routine postoperative antiplatelet therapy was not administered in the Coverage Group or in the patients without stent implantation.

Continuous variables were expressed as mean $\pm$ standard deviation for normally distributed data or as median (Q1, Q3) for non-normally distributed data, and were compared between groups using Student’s t-test or the Mann-Whitney U test, as appropriate. Categorical variables were presented as frequencies and percentages, and group comparisons were performed using Fisher’s exact test or the chi-square test. All p-values were two-sided, with statistical significance defined as p 0.05. Statistical analyses were conducted using SPSS version 26.0 (IBM Corp., Armonk, NY, USA).

Patients were followed up through outpatient visits and telephone interviews. The follow-up duration ranged from 1 month to 13 years. All patients were recommended to undergo CTA prior to discharge and annually thereafter, or earlier if clinically indicated, to monitor for complications such as endoleaks, stent migration, or changes in vertebral artery patency.

A total of 56 patients with aortic arch disease and ILVA who underwent TEVAR were included in this study (Fig. 3). Patients were divided into either the Coverage Group (N = 23) or the Fenestration Group (N = 33), depending on whether fenestration was performed (Table 1). The mean age of the overall cohort was 54.48 $\pm$ 10.31 years, with 89.29% male, and the mean body mass index was 25.88 $\pm$ 3.5 kg/m². The primary diagnoses were acute aortic syndrome (55.36%) and aortic ulcer (21.43%), and chest pain was the most frequent presenting symptom (64.29%). Patients in the Fenestration Group were significantly older than those in the Coverage Group (56.82 $\pm$ 8.78 vs. 51.13 $\pm$ 11.56, p = 0.04). No significant differences were observed between the two groups with respect to gender, presenting symptoms, medical history, or underlying diagnosis. Preoperative CTA showed comparable left vertebral artery (LVA) diameters between the Coverage and Fenestration Groups (3.07 $\pm$ 0.69 mm vs. 3.06 $\pm$ 0.64 mm, p = 0.97), as well as comparable right vertebral artery (RVA) diameters (3.85 $\pm$ 0.69 mm vs. 3.66 $\pm$ 0.40 mm, p = 0.25).

Fig. 3.

Patient selection and study flowchart. CTA, computed tomography angiography.

The procedural outcomes are detailed in Table 2. The overall technical success rate of TEVAR was 98.21%, and no perioperative mortality occurred. A significantly higher proportion of patients in the Fenestration Group underwent simultaneous left subclavian artery stenting compared with the Coverage Group (57.58% vs. 21.74%, p = 0.008). No significant differences were observed between the two groups with respect to endoleak rates, perioperative complications, or postoperative length of stay.

A total of 54 patients (96.43%) underwent CTA at discharge. Compared with preoperative and postoperative measurements, the Coverage Group exhibited a significantly greater reduction in LVA diameter at discharge than that of the Fenestration Group 13.64% [5.52, 22.4%] vs. 0 [–3.29%, 5.13%], p 0.001). The RVA diameter increased by 2.56% in the Coverage Group, whereas the Fenestration Group demonstrated a 2.38% decrease. Although LVA occlusion rates at discharge were similar between groups (8.7% vs. 0%, p = 0.16), the incidence of LVA diameter reduction was significantly higher in the Coverage Group than in the Fenestration Group (39.13% vs. 6.06%, p 0.01).

Among the 38 patients who underwent follow-up CTA, the median follow-up CTA interval was 41.38 (21.42, 73.33) months. The follow-up CTA interval was significantly longer in the Coverage Group than in the Fenestration Group (79.33 [34.25, 106.73] vs. 36.43 [19.67, 55.67] months, p = 0.010). The incidence of complications during follow-up was comparable between groups (17.39% vs. 9.09%, p = 0.61). In the Fenestration Group, one patient experienced an ischemic stroke and two patients developed endoleaks on follow-up CTA. In the Coverage Group, adverse events included one sudden unexplained death, one type II endoleak from the LSA that was treated with coil embolization at 9 months, one reintervention with repeat TEVAR for a new distal aortic ulcer, and one case of LSA stent occlusion. Additionally, a type III endoleak occurred in one Coverage Group patient, which resolved spontaneously at the 3-month CTA follow-up. All endoleaks were either self-limiting or successfully treated, and no long-term adverse events were reported. No patient in the Coverage Group required carotid-subclavian bypass surgery for posterior circulation hypoperfusion or upper limb ischemia.

A total of 38 patients underwent follow-up CTA, including 17 in the Coverage Group and 21 in the Fenestration Group, with a mean interval of 23.29 $\pm$ 32.09 months , see Table 3. At the final CTA, the LVA diameter was significantly greater in the Fenestration Group than in the Coverage Group (2.93 $\pm$ 0.79 mm vs. 1.78 $\pm$ 1.40 mm, p = 0.006), while the RVA diameter was larger in the Coverage Group (4.20 $\pm$ 0.60 mm vs. 3.62 $\pm$ 0.53 mm, p = 0.003). Compared with the preoperative values, the degree of LVA diameter reduction at follow-up was significantly greater in the Coverage Group (28.57% [10%, 100%]) than in the Fenestration Group (0 [–3.03%, 8.33%], p = 0.005). The incidence of LVA diameter reduction was also higher in the Coverage Group than in the Fenetration Group (52.94% vs. 14.29%, p = 0.020). However, LVA occlusion rates did not differ significantly between groups (29.41% vs. 4.76%, p = 0.070). A representative case illustrating ILVA diameter reduction and absence of contrast opacification of the LSA at four-year follow-up is shown in Fig. 4.

Fig. 4.

Preoperative and postoperative 3D-CTA images of a patient in the Coverage Group. (A) Preoperative image showing a prominent ILVA arising directly from the aortic arch, with normal caliber and course. (B) Four-year postoperative image showing a marked reduction in ILVA diameter, indicating chronic hypoperfusion, and absence of contrast opacification in the LSA. S, superior; I, inferior.

This study has several limitations. First, being a retrospective, single-center study, inherent biases related to patient selection, data collection, and analysis may affect the generalizability of our findings. For instance, patients in the Fenestration Group were significantly older, which may reflect a clinical preference for preserving posterior circulation in older patients who have more comorbidities, thus introducing potential selection bias. Second, the relatively small sample size may have limited the statistical power to detect subtle differences between the coverage and fenestration groups. Third, although our study focused on the hemodynamic impact of ILVA preservation, neurological events were not predefined as primary endpoints, and no postoperative neuroimaging, such as brain MRI, was routinely performed. Finally, long-term functional outcomes, including cerebrovascular events and patient-reported quality of life, were not evaluated. These additional endpoints may provide further insights into the clinical significance of ILVA revascularization.