CASE CONFERENCE Frederick A. Henslqy, JI; MD Solomon Aronson, MD, FACC Section Editors CASE 2-2000 Tkansesophageal Echocardiography-Associated Gastrointestinal Trauma Ian Kallmeyer, MB, ChB, David S. Morse, MD, Simon C. Body, MB, ChB, and Charles D. Collard, MD Case Presentation An Sl-year-old woman presented for combined elective coronary artery bypass graft surgery and aortic and mitral valve replacement. Her past medical history included aortic and mitral stenoses, chronic stable angina, congestive heart failure, hypo- thyroidism, breast cancer, and Zenker’s diverticulum that was surgically corrected in 1994. A preoperative transthoracic echocardiogram demonstrated a left ventricular ejection frac- tion of 65%, severe concentric left ventricular hypertrophy, and stenoses of the aortic (valve area, 1.0 cm2; peak gradient, 67 mmHg) and mitral (valve area, 1.6 cm2; peak gradient, 6 mmHg) valves. Additionally, cardiac catheterization revealed multivessel coronary artery disease, including stenoses of the right coronary (50%), first diagonal (600/o), ramus (60%), and acute marginal (100%) arteries. Preoperative chest radiograph and laboratory values were within normal limits. Since surgical repair of Zenker’s diverticulum in 1994, the patient denied any further symptoms of dysphagia, odynopha- gia, or gastroesophageal reflux, and no further gastroenterology consultation, radiologic studies, or endoscopies had been per- formed. In light of the patient’s significant valvular disease and because the patient was without symptoms attributable to residual esophageal disease, consent was obtained for the patient to undergo intraoperative transesophageal echocardio- graphy P’W. After premeditation with intravenous diazepam, 0.1 mg/kg, and fentanyl, 0.5 pgikg, general anesthesia was induced with fentanyl, 15 @kg, etomidate, 0.2 mg/kg, and pancuronium, 0.1 mglkg. After atraumatically intubating the patient with an S-mm endotracheal tube, a multiplane TEE probe (Acuson, Mountain View, CA) was easily passed into the distal esophagus (30 cm at the teeth) on the first attempt using direct visualization of hypopharynx with a laryngoscope blade. TEE with high-quality imaging was then performed’without apparent incident before From the Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA. Address reprint requests to Charles D. Collard, MD, Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women S Hospital, Harvard Medical School, 75 Francis Street, Boston. MA 02115. Copyright 0 2000 by U!B. Saunders Company 1053-0770/00/1402-0023$10.00/O doi:10.1053/c~2000.4688 Key words: gastrointestinal hemorrhage, complications, review, risk factors, cardiac surgery and after cardiopulmonary bypass (CPB). At no point during the case was the probe manipulated while in a locked or flexed position. During CPB (180 minutes), the probe was left in a neutral position with the acoustic power off. The patient was successfully weaned without difficulty from CPB on intrave- nous dopamine, 2 pglkgfmin, and remained hemodynamically stable throughout the remaining intraoperative period (ie, left ventricular function was preserved as determined by TEE). The hematocrit immediately after CPB was 26%, and the patient was transfused with 2 units of packed red blood cells (PRBCs). At the conclusion of the case, bright red blood was noted on the TEE probe when it was removed from the esophagus. After placement of an orogastric tube, 200 mL of bright red blood mixed with coffee grounds was suctioned from the gastrointesti- nal tract. Because the patient was hemodynamically stable and the prothrombin time, partial thromboplastin time, and Intema- tional Normalized Ratio (INR) were within normal limits, the patient was transferred to the intensive care unit (ICU) on intravenous dopamine, 2 yglkglmin, and propofol, 20 ug/kgl min. On arrival in the ICU, however, the patient became hypotensive (arterial blood pressure, 88/47 mmHg; central venous pressure, 2 mmHg; pulmonary artery pressure, 25/5 mmHg), and bright red blood (800 mL) continued to be suctioned from the orogastric tube. The patient’s hematocrit was 25% despite minimal chest tube drainage (100 mL) and having received 2 units of PRBCs. Esophagogastroduodenoscopy (EGD) and a chest radiograph were performed to identify the source of the gastrointestinal bleeding and to rule out the presence of mediastinal air. Although the chest radiograph was within normal limits, EGD revealed large blood clots in the hypopharynx and esophagus associated with several linear esophageal erosions. A large contusion and mucosal tear at the gastroesophageal junction were also noted. There was no evidence of a residual esophageal diverticulum or hiatal hernia. A large amount of blood could be seen trickling from the esophagus into the stomach on a retroflexed view. No injury to the stomach or proximal duode- num was seen. A diagnosis of esophageal contusion and Mallory-Weiss tear was made. The patient was conservatively managed with antihistamine receptor blocker (Hz blocker) therapy, removal of the orogastric tube, and frequent checks of the hematocrit. Systemic anticoagu- lation therapy after mechanical valve replacement was post- poned for 1 week. On this conservative regimen, the esophageal bleeding resolved spontaneously, with the patient requiring 3 212 Journal of Csrdiothoracicand VascularAnesthesia, Vol 14, No 2 (April), 2000: pp 212-216