In Response: We thank Drs. Mackenzie and Asai for their comments on our case report. We feel that Dr. Mackenzie's proposal, to add additional air to the tracheal cuff in an effort to centralize the tracheal lumen of the double-lumen endotracheal tube, is a good idea and should be tried in a similar situation. It would seem prudent to monitor cuff pressure if this is attempted. We had filled our tracheal cuff with air via the minimal leak technique and had no evidence of a tracheal cuff leak at the time of the obstruction. We agree that segments of U-shaped tracheas as defined by Dr. Mackenzie are common, and that considerable variability of shape is also common within each trachea [1,2]. The "saber-sheath" trachea as defined by Greene and Lechner [3] had an internal coronal diameter one-half or less of the corresponding sagittal diameter, and this finding occurred over the length of the intrathoracic trachea. Our case report met this definition on review of our posteroanterior and lateral chest radiographs and at multiple levels of the intrathoracic trachea on chest computed axial tomographs. We apologize if omission of this information caused confusion. In people without chronic obstructive pulmonary disease (COPD), saber-sheath trachea is clearly rare [4]. It seems to be present primarily in a small percentage of older men with COPD [3,5]. We agree with Dr. Mackenzie's assertion that chest computed axial tomographs will give a more accurate representation of the shape and variations in diameter of the trachea than conventional chest radiographs. Our point, and that of Dr. Asai, was that clinically significant narrowing of segments of the trachea can frequently be identified by routine chest radiographs, and this potential should not be overlooked [6]. In response to Dr. Asai's letter, the patient had no complication postthoracotomy attributable to the endotracheal tubes. His doublelumen tube was changed to a single-lumen 7.0 Fr tube at the end of the case uneventfully. He was extubated on the third postoperative day; chest tubes were removed by the fourth postoperative day without a persistent pneumothorax. We do not agree with Dr. Asai's implication that intubation in patients with "saber-sheath" trachea should be avoided if general anesthesia is required. To date, there are no reports of tracheal damage from intubation in these patients. We are concerned that the alternatives of mask of laryngeal mask airway ventilation may not adequately ventilate some of these patients, considering that the majority of them have COPD. We feel that control of the airway by intubation with a conventional single-lumen endotracheal tube, having an external diameter shorter than the most narrow diameter of the trachea, if clinically indicated, is appropriate. Joseph Bayes, MD Eliot M. Slater, MD Peter S. Hedberg, MD Dexter Lawson, MD Department of Anesthesiology, Salem Hospital, Salem, MA 01970-2768