Resection and replacement of thoracic aortic graft infections.

Abstract

Thoracic aortic graft infection (TAGI) presents a formidable challenge with high mortality. We evaluated our 22-year experience managing TAGI with extensive debridement, graft replacement, vascularized tissue coverage, and aggressive antibiotics. We reviewed all consecutive patients with TAGI from 1991 to 2013. We also compared infected cases versus noninfected reoperative controls using a case-control design. Standard statistical methods were used for descriptive analysis, and Kaplan-Meier for survival analysis. We treated 32 TAGI patients, involving 19 ascending/arch (A/A) and 13 descending/thoracoabdominal (D/TAA) grafts, including 4 endografts. In total, 19 (59.4%) presented with pseudoaneurysm and 11 (34.4%) with aortic fistula. Vascularized tissue (omentum or muscle) coverage was possible in 22 (71.0%) patients. Thirty-day mortality occurred in 3 (9.4%) patients, with no 30-day mortality among those receiving vascularized graft coverage (P = .018). During follow-up, reinfection occurred in 8 patients (25% [4 A/A and 4 D/TAA]). Five-year overall (A/A 45.4% vs D/TAA 28.9%, P = .434) and reinfection-free (A/A 19.2%, D/TAA 27%, P = .409) survival was similar between groups. Long-term mortality was greater after endograft infection (100% vs 25% at 2.5 months, P = .0007) or aortobronchial fistulization (100% vs 37.9% at 6 months, P = .026). Time to reintervention was shorter in infected versus non-infected reoperative cases (31 vs 83 months, P < .0001), but there were no significant differences in long-term mortality after reoperation. TAGI continues to represent a highly morbid surgical challenge. Prompt antimicrobial coverage, debridement, graft replacement, and vascularized graft coverage, yielded best long-term results. Endograft infection and aortobronchial fistula had very poor prognoses.