Recovery Takes Time: Loss of Signal Predicts Delayed Recovery of Vocal Cord Function After Thyroidectomy

Background: Post-thyroidectomy vocal cord dysfunction (PT-VCD) is an important side effect of thyroid surgery. With the introduction of intraoperative neuromonitoring (IONM), hopes have been raised that either the rate or severity of PT-VCD could be reduced. However, data to support these concepts are scarce. To better understand the relationship between IONM outcomes and the severity of PT-VCD, a detailed time-course evaluation of recovery of PT-VCD was performed in a continuous clinical quality registry from a specialized high-volume endocrine surgery center. Methods: Data were prospectively recorded in a single-center clinical quality assurance registry (June 2015 to May 2016) and subsequently analyzed retrospectively, with a 12-month follow-up for all cases. All patients underwent vocal cord (VC) laryngoscopy (VCL) by independent ear-nose-throat (ENT) specialists before and after surgery. Cases with newly diagnosed PT-VCD were enrolled in a detailed follow-up program (recruitment from June 2015 to May 2016) that included structured telephone interviews every 4-6 weeks to assess the exact time course of PT-VCD recovery and VC status for a period of at least 12 months. Clinical data were analyzed for variables affecting the time course of recovery by univariate analysis. Results: From 6/2015 to 5/2016 there were 1097 consecutive thyroid procedures. During this period, 78 cases of PT-VCD (1591 nerves at risk (NARs); 4.9%) were entered into the detailed follow-up-program. Of these, three cases of PT-VCD persisted at 12 months (PT-VCD 0.18% NAR), with six cases lost to follow-up (maximum rate of potentially persistent PT-VCD of 0.54% NAR). In total, 15% of PT-VCD cases recovered within 4 weeks; the mean recovery time was 4.4 months, and 6 months after thyroidectomy, 18% still had impaired VCL tests. Individual cases were followed >12 months showing late full recovery of PT-VCD, thereby challenging the definition of permanent VCD. Logistic regression analysis revealed non-transitory loss of signal (ntLOS) (OR for recovery within 12 weeks: 0.39; 95%CI 0.15-0.98; p = 0.046) and more specifically, secondary ntLOS, to be a significant independent predictor of PT-VCP recovery beyond 12 weeks (OR for recovery within 12 weeks 0.303; 95%CI 0.115-0.797; p = 0.016). Conclusions: For the first time, these data provide a detailed description of the time course of PT-VCD recovery in a large cohort, along with correlations to operative data and IONM findings. Our study indicates that recovery from PT-VCD can be prolonged, and specifically, the occurrence of ntLOS-especially secondary ntLOS-during IONM was predictive of a longer recovery trajectory. This suggests that IONM may offer an additional advantage by functioning as a prognostic tool, helping to identify patients at higher risk for extended recovery periods. Such early identification could enable a more targeted approach, potentially allowing for the earlier initiation of supportive interventions, like speech therapy, in those most likely to benefit from proactive management.