Mycotic thoracic aortic aneurysm (MTAA) is a disease that is difficult to treat and often lethal. Open repair has high morbidity and mortality risks; additionally, thoracic endovascular aneurysm repair (TEVAR) often requires innovative techniques.

We report the use of an innominate artery chimney endovascular aneurysm repair (ChEVAR) in conjunction with an open carotid-carotid and carotid-left subclavian artery (LSA) bypass to treat a time-sensitive Salmonella-related mycotic thoracic aortic aneurysm (MTAA). A symptomatic early post-operative type 1a endoleak was also successfully treated.

The study was approved by the Institutional Review Board of the Singapore General Hospital (IRB no. 2023/2291).

A 68-year-old male with multiple comorbidities, including hypertension, ischemic heart disease, and an ejection fraction of 30%, underwent cervical spine laminectomy and fusion for severe spondylosis with spinal stenosis and myelopathy. The patient had previously received multiple percutaneous coronary interventions.

Two weeks after surgery, he presented with septic shock, with symptoms such as a temperature of 38.8°C, chills, and rigor. He did not experience severe chest pain and gastrointestinal symptoms. Although he had no recent travel history, he had consumed some raw produce. The recent surgical wounds appeared benign.

While cardiac investigations did not yield significant results, inflammatory markers were notably elevated: serum C-reactive protein level was 218 mg/L, erythrocyte sedimentation rate was 80 mm/h, and white cell count was 22 cells/μL.

The patient responded to early fluid resuscitation, broad-spectrum intravenous (IV) antibiotics, and low-dose inotropes. Subsequent abdominal imaging and magnetic resonance imaging of the spine revealed no significant pathology or peri-prosthetic collections. Two-dimensional echocardiography showed no cardiac vegetations.

Positron emission tomography (PET)-computed tomography (CT) was then performed to identify the source of sepsis, revealing fluorodeoxyglucose avid saccular thoracic aneurysms. Further characterization using a CT thoracic aortogram identified two contrasting and possibly mycotic outpouchings distal to the LSA that measured 3 and 0.8 cm, respectively (Fig. 1A, B).

Figure 1. (A, B) Computed tomography (CT) aortogram showed outpouchings that indicate a mycotic aneurysm (^), (*). (C) Repeated CT aortogram revealed the increase in size of outpouchings (approximately 5 mm).

Blood cultures were positive for Salmonella Enteritidis, for which the patient was treated with culture-directed IV ceftriaxone. The disease was further managed with the assistance of an infectious disease physician. The patient was initially scheduled for thoracic endovascular aneurysm repair (TEVAR) after a duration of ‘sterilization’ with broad-spectrum IV antibiotics; however, he complained of increasing dull chest pain on the third day of admission. A repeat CT aortogram indicated that the larger outpouching had increased by 5 mm to reach 3.5 cm in diameter (Fig. 1C).

For treatment, a hybrid surgical approach was chosen, consisting of open carotid-carotid and carotid-LSA bypasses, complemented by endovascular aortic arch stenting with an innominate chimney (Fig. 2). This operation followed a series of seronegative blood cultures.

Figure 2. (A) Schematic diagrams showing aortic arch with saccular outpouchings. (B) Planned repair with 46-42 thoracic stent grafts, innominate chimney, and bypasses, as well as AVP plug and ligated left common carotid artery. AVP, Amplatzer Vascular Plug.

For stent graft sizing, the ascending aorta, proximal to the innominate artery, measured approximately 36.5 mm. A 16-mm BeGraft (Bentley InnoMed GmbH) was selected for the innominate chimney. Considering high pressure, location, and necessary oversizing, the minimal stent diameter had to be 42 mm with an additional 5 mm to wrap around the chimney (1/3 of 16 mm), as recommended by the manufacturer. Consequently, the largest available option selected was a tapered 46-42 mm Cook Zenith Alpha thoracic aortic stent graft. The diameter of the descending aorta measured approximately 31 mm. For this, 150- and 200-mm stent grafts were available and smaller- diameter (34-36 mm) distal extension stent grafts were considered but ultimately not selected, primarily due to cost considerations (Fig. 2).

Intraoperatively, two longitudinal incisions were made along the anterior border of each sternocleidomastoid muscle to expose the right and left carotid arteries, with a third transverse incision in the left supraclavicular fossa to expose the LSA. Ring-reinforced, rifampicin-soaked GORE Propaten grafts of 8 mm were gently tunneled between the incisions: one behind the esophagus for the carotid-carotid bypass and another between the left longitudinal and left supraclavicular incisions. Graft-to-vessel, end-to-side anastomoses were performed following the standard procedure.

The carotid-carotid bypass was performed distally on the medial sides of the right and left carotid arteries, whereas the left carotid-to-subclavian bypass was performed more proximally from the lateral side of the left carotid artery to the subclavian artery, resulting in a step-ladder-type configuration from right to left.

Bilateral common femoral arteries were accessed percutaneously, and an ascending aortogram was performed, revealing saccular outpouchings at Ishimaru zone 3 (Fig. 3A) [1]. A 14-mm Amplatzer Vascular Plug (AVP) II (Abbott) was placed in the LSA proximal to its vertebral artery origin, before positioning the main body of a 46-42-179 Cook Zenith Alpha Thoracic Endovascular Stent graft over a Lunderquist wire proximal to the innominate artery origin (Fig. 3B). Next, the right axillary artery was accessed percutaneously for cannulation and stenting of the innominate artery using a long 12 Fr sheath. The main thoracic stent graft and two innominate BeGrafts (16×58 mm and 16×38 mm) were placed and balloon-molded simultaneously.

Figure 3. (A) Initial aortogram showed the mycotic thoracic aortic aneurysm, left common carotid artery (LCA)-left subclavian artery (LSA) bypass (*) with the carotid-carotid bypass out of frame. (B) Amplatzer Vascular Plug II vascular plug (14 mm) placed proximally to the original in the left vertebral artery. (C) Final angiogram demonstrated proximal LSA plug, thoracic stent graft, innominate artery chimney, and both carotid-carotid and LCA to LSA bypasses.

Finally, the origin of the left carotid artery was surgically ligated using the same incision, and a completion angiogram was performed. This showed no endoleak, with contrast opacification of the innominate artery, and well patent left carotid and LSA via the bypasses (Fig. 3C).

Post-operatively, the patient remained stable and recovered without neurological complications. However, on the fourth post-operative day, he complained of persistent chest pain accompanied by persistent sac filling proximal to the LSA, suggestive of a type 1a gutter endoleak, as identified on a CT scan (Fig. 4).

Figure 4. Computed tomography aortogram revealed a likely gutter type 1a endoleak.

The patient underwent embolization of the gutter endoleak through left brachial access, moving past the AVP plug of the LSA using a Bernstein catheter and Terumo 035 wire. This approach allowed the placement of 18 Ruby coils from Penumbra Inc. into the aortic space (Fig. 5). Embolic glue was not used and the left vertebral artery remained patent after the procedure.

Figure 5. (A) Angiogram through left brachial access demonstrating endoleak and an artery of unknown significance (*). (B) Placement of 18 Ruby coils to correct the endoleak.

The chest pain improved shortly thereafter, and the patient was discharged on lifelong oral ciprofloxacin. Outpatient CT aortograms performed at 6 months and 1 year after surgery showed resolution of the endoleak with no increase in aneurysm sac size. The bypasses remained patent, and the patient has continued to maintain good health.

MTAA, defined as localized dilatation of the aorta due to infection-induced weakening of the vessel wall, presents a significant surgical challenge and carries a high risk of lethality.

Four mechanisms of pathogenesis have been proposed: bacterial wedging at sites of the existing intimal injury or atherosclerotic plaque, septic emboli into the arterial vasa vasorum, extension of a contiguous infective focus into the aortic wall, and direct bacterial inoculation during penetrating trauma [2].

Gram-positive bacteria, including Staphylococcus, Enterococcus, and Streptococcus spp. account for 60% of causative organisms, while Salmonella app. account for the majority of infections caused by gram-negative bacteria. Mycobacterium tuberculosis and Treponema pallidum are also noted causative organisms [3].

Salmonella spp. have an inclination for adherence to diseased vascular endothelium [4], and non-typhoidal Salmonella spp. constitute the majority of cases in East Asia [5,6], where infection often develops after the consumption of uncooked eggs and meat. Although Salmonella-related mycotic aortic aneurysm tend to progress quickly and with higher rupture rates, effective antimicrobial control has led to favorable long-term outcomes and infection-related complications, as demonstrated in one of the largest case series by Sörelius et al. [7].

Aortic arch anatomy presents challenges for TEVAR because of the proximity of the aortic valves, presence of supra-aortic branches [8], and the need to consider both inner and outer curvatures when selecting stent size. Although various techniques, including fenestrated or custom-made devices, have increased the TEVAR armamentarium in the elective setting, their extensive manufacturing time limits use in time-sensitive cases.

Open repair is associated with high morbidity and mortality, rendering it unsuitable for treating frail patients with multiple cardiovascular comorbidities. While there is no gold standard treatment for MTAAs, TEVAR has been extensively studied since 1998 [9], with multiple subsequent case series [10-12] documenting comparable short- and long-term survival rates to open repair [13].

In urgent situations, chimney grafts or stents placed parallel to the main aortic stent graft can preserve flow to essential branches and are effective when correctly sized. Initially described as rescue measures for inadvertent coverage of visceral vessels during standard endovascular aneurysm repair (EVAR) [14], chimney grafts have demonstrated favorable short- and mid-term outcomes in time-sensitive or salvage para-visceral EVAR procedures [15]. However, the use of chimney grafts in the innominate artery as part of a hybrid approach is seldom reported, even in some of the largest series of MTAAs.

Although the patient in this report developed an early symptomatic type 1a endoleak, we treated the condition successfully. However, this case underscores the need for further research into not only prophylactic gutter embolization but also optimal stent graft oversizing to enhance the treatment efficacy [16,17]. When failing to navigate past the AVP in the LSA from the left brachial artery, access may be obtained through the aortic sac via the left carotid stump, allowing cannulation to embolize the sac. The high risk of a type I endoleak with ChTEVAR has been consistently reported, including a report by Kanaoka et al. [18], where 48.5% of their 33 single-ChEVAR cases experienced endoleaks. Therefore, we hypothesize that prophylactic gutter embolization may be a possible option during initial treatment because of the high prevalence of type 1 endoleak.

In summary, hybrid TEVAR with an innominate artery chimney and supra-aortic open bypasses is a feasible option for time-sensitive cases. Despite an early and significant endoleak, the treatment procedure was performed safely. The efficacy of prophylactic gutter embolization remains to be fully understood; therefore, further studies are required.

None.

The authors have nothing to disclose.

Concept and design: NNZP, SPYD, LQS. Analysis and interpretation: NNZP. Data collection: LQS. Writing the article: SPYD, NNZP. Critical revision of the article: NNZP, LQS. Final approval of the article: all authors. Statistical analysis: none. Obtained funding: none. Overall responsibility: LQS.