Intra- and Extrathoracic Malignant Tracheoesophageal Fistula-A Differentiated Reconstructive Algorithm

Tracheoesophageal fistulae (TEF) after oncologic resections and multimodal treatment are life-threatening and surgically challenging. Radiation and prior procedures hamper wound healing and lead to high complication rates. We present an interdisciplinary algorithm for the treatment of TEF derived from the therapy of consecutive patients.

18 patients (3 females, 15 males) treated for TEF from January 2015 to July 2017 were included. Two patients were treated palliatively, whereas reconstructions were attempted in 16 cases undergoing 24 procedures. Discontinuity resection and secondary gastric pull-up were performed in two patients. Pedicled reconstructions were pectoralis major (n = 2), sternocleidomastoid muscle (n = 2), latissimus dorsi (n = 1) or intercostal muscle (ICM, n = 7) flaps. Free flaps were anterolateral thigh (ALT, n = 4), combined anterolateral thigh/anteromedial thigh (ALT/AMT, n = 1), jejunum (n = 3) or combined ALT-jejunum flaps (n = 2).

Regarding all 18 patients, 11 of 16 reconstructive attempts were primarily successful (61%), whereas long-term success after multiple procedures was possible in 83% (n = 15). The 30-day survival was 89%. Derived from the experience, patients were divided into three subgroups (extrathoracic, cervicothoracic, intrathroracic TEF) and a treatment algorithm was developed. Primary reconstructions for extra- and cervicothoracic TEF were pedicled flaps, whereas free flaps were used in recurrent or persistent cases. Pedicled ICM flaps were mostly used for intrathoracic TEF.

TEF after multimodal tumor treatment require concerted interdisciplinary efforts for successful reconstruction. We describe a differentiated reconstructive approach including multiple reconstructive techniques from pedicled to chimeric ALT/jejunum flaps. Hereby, successful reconstructions are mostly possible. However, disease and patient-specific morbidity has to be anticipated and requires further interdisciplinary management.