Improving the outcome of esophagectomy : anatomic, surgical and postoperative aspects

Abstract

Esophagectomy is known for its high postoperative morbidity and mortality rates. In case of esophageal cancer, a transthoracic approach is often preferred since this allows for mediastinal lymphadenectomy, removing potential metastases. However this is frequently associated with pulmonary complications, increasing the need for intensive care unit treatment and mortality. Therefore this thesis aims to improve the outcome of esophagectomy. A literature review shows that pulmonary complications are reduced by a minimally invasive approach and starting enteral nutrition early postoperatively. Preoperative optimization of patient performance, administration of anti-inflammatory medication perioperatively and prone-positioning are promising interventions, requiring further investigation. However, valid inter-study comparisons cannot be made since many different definitions are used for pneumonia. Therefore a recently proposed definition, the Utrecht Pneumonia Score (UPS) was revised and internally and externally validated, showing excellent discrimination and calibration. Currently this is the only validated definition for post-esophagectomy pneumonia available. Inflammation may be the final common pathway in the multifactorial etiology of pulmonary complications following esophagectomy. Consequently several randomized studies investigated the effect of perioperative glucocorticoid administration, but sample sizes were too small. Therefore, we performed a meta-analysis, showing that weight dependent dose of methylprednisolone significantly reduces pulmonary complications, emphasizing the hypothesis that inflammation may be pivotal. Another important regulatory pathway in inflammation and pulmonary function, affected by esophagectomy, runs via the vagus nerve. A vagotomy is performed as part of an esophagectomy due to its course next to the esophagus, but selective sparing of pulmonary branches may be feasible. Since the regional anatomy is complex we performed an anatomical and histological study, a study during thoracoscopic esophagectomies in vivo, a magnetic resonance imaging study in vivo. These show extensive pulmonary innervation, and indicate how these vagal branches can be spared. These also describe the meso-esophagus, a previously undescribed connective tissue layer. This courses from esophagus to aorta, dividing the posterior mediastinum in two compartments. Based on the anatomical mapping, feasibility of thoracoscopic pulmonary vagus nerve branches sparing esophagectomy was studied in human cadavers. This procedure is feasible, but attention should be given to removal of all peribronchial lymph nodes. The vagus nerve can be activated through enteral nutrition leading to dampening of postoperative inflammation. However, a nil-by-mouth strategy following esophagectomy is standard of care in many centers. Through a systematic literature review and retrospective evaluation of routine jejunostomy tube feeding, we demonstrated the need to investigate early oral intake. However, concerns are increased severity of anastomotic leakage and (aspiration) pneumonia. Therefore we performed a single-arm, multicenter trial showing that early oral intake following esophagectomy is feasible and does not increase morbidity. Chyle leakage is a complication of enteral nutrition. Through a large retrospective analysis we have shown that the majority can be managed by dietary measures only. In conclusion in this thesis a technique to spare the pulmonary branches of the vagus nerve during thoracoscopic esophagectomy is validated and the safety and feasibility of early start of oral intake following esophagectomy are shown. Further studies are needed to determine the effects of these interventions.