A systematic review of therapies for aortobronchial fistulae.

Objective: The aim of the study was to summarize epidemiologic data about aortobronchial fistulas and compare outcomes (mortality, recurrence, re-operation) of open, staged, and endovascular repair of aortobronchial fistula.

Methods: A systematic literature review was conducted to identify eligible studies published between January of 1999 and December of 2019. The Cochrane Library, PubMed and Scopus databases were used as search engines. Eligible studies included articles reporting postoperative outcomes (death/follow-up). Literature review revealed only case reports and small case series and thus, only descriptive data with data heterogeneity was available. The corresponding authors were contacted to provide additional information or outcome updates (recurrence/reoperation/death).

Results: Overall, 214 patients (90 studies) underwent 271 procedures (including re-do procedures and staged procedures). Most of the patients were treated by endovascular means (72.42%). Open surgical repair was performed in 21.96% and staged procedures in 5.6%. Aortobronchial fistulae located most often in the descending thoracic aorta (Zone 3,4) (64,6%) and in Zone 2 (23,8%). Fourteen percent of aortobronchial fistulae developed after thoracic endovascular aneurysm repair. Recurrence or infection occurred in 20% (43 patients). Recurrences were at some extend associated with the presence of endoleak. Long-term antibiotic administration (>1 month) was instituted in 63 patients (29.4%), whilst 90 patients (42%) did not receive antibiotics beyond hospitalization. From the remaining 61, 3 received life-long antibiotics and for 58 patients data were not available. Considering outcomes, mean follow-up was 25.1 months (0-188 months) and not significantly different among treatments.

Limitations: Literature review has revealed only case reports and small case series and thus, only descriptive data were available. Randomized controlled trials are not available due to the rarity of the disease which significantly decreases the power of the present study. Also, this study reflects significant data heterogeneity due to the nature of the analyzed manuscripts and would benefit from large patient cohort studies which till today have not been conducted.

Conclusion: Aortobronchial fistula is a complex disease. Endoleaks may be involved in the development and in recurrence process and they should not be disregarded. Considering major outcomes (length of follow-up), the available treating strategies are equal and thus, surgeons should feel confident to apply the treatment of their choice, taking in mind their experience, patient's age, and clinical condition.