Preservation of the infected thoracic aortic endograft with thoracoscopic drainage and continuous irrigation

The patient was a 72-year-old man, whose previous surgical history included: intestinal repair for its rupture resulting from a traffic accident at 40 years of age, cholecystectomy for choledocholithiasis at 62 years, and pancreatoduodenectomy for pancreatic neoplasms when he was 63 years old. Although these previous abdominal surgeries had provoked recurrent cholangitis, his cholangitis had been treated well and under control in those days. We subsequently considered the intervention for thoracic aortic aneurysm, and presented him with the surgical choices: aortic surgery or thoracic endovascular aortic repair (TEVAR). He chose TEVAR because of its less invasiveness, and then we proceeded with TEVAR, with right common carotid–left common carotid–left subclavian artery bypass (Fig. 1a). He recovered well and was discharged without any complication. One month after the TEVAR, he was readmitted with high-grade fever and diagnosed with a recurrence of cholangitis. Escherichia coli was detected in his blood culture. Computed tomography (CT) on admission did not demonstrate endograft infection (Fig. 1b). Although the patient was started on cefozopran (3.0 g/day), he remained febrile (38.0–38.5 °C); his white blood cell (WBC) count was 15,000–20,000/µL and C-reactive protein (CRP) level was 15–20 mg/dL for more than 7 days. Repeated CT revealed contrast enhancement around the endograft, fluid collection inside the remnant aneurysm (Fig. 1c-1) and a fluid cavity (roughly the size of a chicken egg) between the aortic arch and pulmonary artery (Fig. 1c-2). No endoleak was confirmed. We assumed that the pathogen of cholangitis spread via blood stream, which resulted in endograft infection. Endograft infection might induce endoleak, endograft migration or aortic rupture. Although his cholangitis might be on active phase, we had priority to treat endograft infection initially.

Though the endograft excision with aortic reconstruction was optimal, we considered that the patient could not tolerate invasive aortic surgery with thoracotomy. His activity of daily life (ADL) was maintained at low level; he could communicate with others properly and did not have cognitive dysfunction such as dementia. However, he needed help for eating, changing clothes, and for the use of toilet, in daily life. He could walk a little but basically was on wheel chair during daytime. He suffered from severe chronic obstructive pulmonary disease, and treated with β2 stimulator inhalation. Pulmonary functional test revealed 45% of forced expiratory volume 1 s % (FEV 1.0%) and 0.9 L of forced expiratory volume 1 s (FEV 1.0). He also had septic cholangitis, with relatively high Euroscore II of 15.6. We also considered replaced-graft reinfection due to active and smoldering cholangitis. In addition, because of his abdominal surgical history, omental coverage was not an option. Therefore, we chose a less invasive and conservative strategy; thoracoscopic drainage with endograft preservation, followed by continuous irrigation.

Using a thoracoscopic surgical system through several access ports, we made an incision on the lateral side of the distal arch aneurysm (Fig. 2a) and created a 5-cm-length fenestration, which extended to the sub-aortic bulging area (Fig. 2b). The purulent fluid, which E. coli was cultured on later, erupted from the fenestration. We performed irrigation with copious saline containing gentamycin, in addition to crystal violet for gram-positive bacteria in case of polymicrobial infection. Two outflow drains (28Fr Argyle™ thoracic catheter and 24Fr Trocar catheter, Covidien, Dublin, Ireland) were placed on the diaphragm and at the apex of the left lung. Two inflow drains (10Fr BLAKE^® Silicone Drains, Ethicon, Newark, USA) (10Fr Nelaton catheter, Terumo, Tokyo, Japan) were placed lying on the fenestration (Fig. 2c).

The irrigation–drainage system was constructed postoperatively. Gentamycin, which is effective in gram-negative rod infection, was empirically diluted in 1000–2000 mL normal saline/day, and did not exceed the intravenous maximum dose (240 mg/day). By postoperative day (POD) 30, the abscesses had diminished upon CT imaging (Fig. 1d). His WBC count was normalized and CRP level stayed less than 0.03 mg/dL. E. coli was not cultured in the drainage fluid and blood, therefore we removed the inflow and outflow chest tubes on POD 32. After confirmation of no recurrent infection following drain removal, we decided to alter an intravenous cefozopran to oral amoxicillin (1.0 g/day) on POD 34. He recovered well and discharged on POD 92 via a rehabilitation hospital. In case of the next step as radical treatment such as total aortic arch replacement, we closely monitored his cholangitis and enhanced his ADL with rehabilitation. Oral amoxicillin was finished on POD 180 with a normal range of WBC count and CRP level. At current follow-up, he is now 28 months after drainage with no sign of recurrent infection including endograft infection and cholangitis (Fig. 1e).