Three-dimensional imaging of the speaking shunt in patients with near-total laryngectomy.

Abstract

Near-total laryngectomy with creation of a speaking shunt is generally considered suitable treatment for most T3 and some T4 laryngopharyngeal cancers. In some patients, poor speech production by the shunt can be problematic and usually means that a shunt lumen is stenotic or too small. Conventional axial computed tomography (CT) of the neck is of little value in predicting the shunt function and the patency of the shunt lumen. Such defect, however, can be better overcome if the spiral CT is used. This study was designed to evaluate the dynamic speaking shunt by spiral CT and a three-dimensional (3-D) imaging model of the upper airway in 59 near-total laryngectomies. During scanning, the patients were asked to make a single, 20-second phonation. A high-quality 3-D surface model was then rendered at an independent workstation. The 3-D model took on the appearance of the speaking shunt which could serve as a tracheopharyngeal fistulography. 3-D reconstruction of the images could clearly demonstrate the dynamic anatomy and patency of the speaking shunt. We found the 3-D model to be of diagnostic value, particularly in the near-total laryngectomy patients with stenotic shunts. The most common site of stenosis was at the top portion of the shunt. The 3-D image findings of the shunts significantly improved the assessment and therapies of the patients undergoing revision surgeries of the shunt stenoses.