Mediastinal fluid as a predictor for esophageal perforation as the cause of pneumomediastinum.

Abstract

Pneumomediastinum is not an uncommon finding on emergency CTs and is seen in a variety of settings, including acute trauma, recent repeated emesis and recent thoracic surgery, and after EGD (Kouritas et al. J Thorac Dis 7(Suppl 1):S44-S49, 2015). The purpose of this study was to examine cases of CT pneumomediastinum for distinguishing features which could support or exclude esophageal injury as a cause, which in turn could help guide evaluation of these patients. CT chest scans showing pneumomediastinum performed in the emergency department between January 2013 and December 2018 were included. The presence or absence of fluid within the mediastinal compartments was correlated with esophageal perforation and subdivided into the clinical scenarios of trauma, suspected Boerhaave's syndrome, recent surgery or EGD, and other. Accuracy of this finding was compared with fluoroscopic esophagography. Twenty-two cases of esophageal perforation were identified out of a total of 324 included cases. The cases were subdivided into four categories: trauma, suspected Boerhaave syndrome, suspected iatrogenic perforation from recent procedure or surgery, and other. Two hundred fourteen cases of pneumomediastinum occurred in the setting of trauma, and 2 had esophageal perforation. Both showed mediastinal fluid. Twenty-two cases had mediastinal fluid without perforation. Seventeen cases of pneumomediastinum occurred in the setting of suspected Boerhaaves, and 3 had esophageal perforation. Every case with esophageal perforation had mediastinal fluid, and every case without perforation had no mediastinal fluid. Nine cases of pneumomediastinum occurred in the setting of suspected iatrogenic perforation after recent surgery or procedure. Six cases had esophageal perforation, and 5 of these had mediastinal fluid. All three cases without perforation also had mediastinal fluid. Eighty-six cases were classified as other and included a variety of clinical histories. This category contained 8 esophageal perforations, 7 of which had mediastinal fluid. One case of mediastinal fluid was not associated with esophageal perforation in this category. The presence of mediastinal fluid, specifically within the visceral compartment, strongly suggests esophageal injury, and its absence strongly argues against it. An important caveat is in the setting of recent surgery, in which mediastinal fluid can be seen normally. An esophagography study can supplement the evaluation if there is a discrepancy between the clinical suspicion and the original CT findings.