Usefulness of intraoperative nerve monitoring in esophageal cancer surgery in predicting recurrent laryngeal nerve palsy and its severity.

Abstract

Recurrent laryngeal nerve (RLN) palsy is a critical postoperative complication in esophageal cancer surgery. However, intraoperative prediction of its occurrence and severity is difficult. In this prospective study, we evaluated the usefulness of intraoperative nerve monitoring (IONM) in predicting RLN palsy and its severity. Twenty patients who underwent subtotal esophagectomy with 3-field lymph node dissection were enrolled. Intraoperative electromyography (EMG) amplitudes of the vocal cords were measured by IONM at RLN and vagus nerve (VN) stimulation. Comparison was made between the vocal cords with RLN palsy and those without palsy and additionally between the vocal cords with transient RLN palsy and those with persistent palsy. Among 40 vocal cords in 20 patients, 26 were intact and 14 were paralyzed. Seven had transient, six had permanent palsy. The mean EMG amplitude of intact vocal cords was significantly larger than that of paralyzed ones at VN (506 ± 498µV vs. 258 ± 226µV, p = 0.022) and RLN stimulation (642 ± 530µV vs. 400 ± 308µV, p = 0.038). The cut-off value for postoperative palsy were 419µV [positive predictive value (PPV): 48.0%, negative predictive value (NPV): 84.6%] at VN and 673µV (PPV: 44.8%, NPV: 90.9%) at RLN stimulation. The mean EMG amplitude of persistently paralyzed vocal cords tended to be small, compared with that of recovered ones at both VN (168 ± 173µV vs. 336 ± 266µV, p = 0.11) and RLN (244 ± 223µV vs. 536 ± 344µV, p = 0.051) stimulation. The absolute EMG amplitude of IONM might be helpful to predict the occurrence and severity of RLN palsy after esophageal surgery although the predictive value is low.