Computed tomography-guided drainage of mediastinal abscesses: clinical experience with 23 patients

To evaluate the technical and clinical success rates of computed tomography (CT)-guided percutaneous drainage of mediastinal abscesses.

An interventional radiology database was used to identify patients who underwent CT-guided percutaneous drainage of mediastinal abscesses. Medical records were reviewed to evaluate abscess etiology, drainage technique, clinical outcome, and complications.

Over a 10-year period, 23 patients (20 men; average age, 54 y; range, 34-77 y) with 24 mediastinal abscesses underwent 25 CT-guided drainage procedures. Abscess etiologies included esophageal leak after esophagectomy (n = 6), perforated esophageal cancer (n = 4), Nissen fundoplication (n = 3), emetogenic esophageal rupture (n = 3), infectious (n = 2), cardiac surgery (n = 1), iatrogenic (n = 1), gastric strangulation (n = 1), Whipple procedure (n = 1), and thoracotomy for lung cancer (n = 1). Drainages were performed with tandem trocar (n = 14) or Seldinger (n = 11) technique. A total of 25 catheters were used: 8.5 F (n = 9), 10 F (n = 8), 12 F (n = 6), 14 F (n = 1), and 16 F (n = 1). The mean time of catheter drainage was 13.6 days. Technical success was achieved in all 25 attempts (100%). Twenty-two of the 23 patients had complete resolution of the abscess without the need for surgical debridement, for a clinical success rate of 95.6%. One patient underwent technically and clinically successful abscess drainage but required surgical exploration for repair of an anastomotic leak after esophagogastrectomy. There was one complication. One patient had inadvertent placement of a catheter within a pulmonary vein. The catheter was removed after 24 hours without hemodynamic consequences.

Percutaneous CT-guided drainage of mediastinal abscesses is an uncommon procedure, but the results of this study suggest that it is associated with high technical and clinical success rates.