The Effect of Diabetic Neuropathy on Recurrent Laryngeal Nerve Monitoring During Thyroid Surgery

TL;DR

Diabetic patients show greater changes in nerve function during thyroid surgery, but these changes don't lead to worse short-term vocal outcomes.

Contribution

First study to quantify RLN amplitude changes during thyroidectomy in diabetic patients using intraoperative nerve monitoring.

Findings

Diabetic patients had significantly greater post-dissection RLN amplitude reduction compared to non-diabetic patients.

Electrophysiological changes in diabetic patients did not correlate with increased short-term vocal fold palsy incidence.

Pre-dissection RLN parameters were similar between diabetic and non-diabetic groups.

Abstract

Background: Recurrent laryngeal nerve (RLN) injury remains a significant complication of thyroid surgery, potentially leading to vocal cord paresis and compromised quality of life. Intraoperative nerve monitoring (IONM) enables real-time assessment of RLN function but may be affected by underlying neuropathies. Diabetes mellitus (DM) induces peripheral nerve damage, potentially altering RLN electrophysiology during surgery. This study investigates the impact of DM on RLN amplitude and latency during thyroidectomy.

Methods: We prospectively analyzed 245 patients (485 RLNs at risk) who underwent thyroidectomy with IONM between 2020 and 2024. Patients were classified as diabetic or non-diabetic according to American Diabetes Association (ADA) criteria. Standardized surgical and anesthetic protocols were followed. RLN amplitude and latency were measured before and after dissection using consistent stimulation parameters, and group comparisons were performed using non-parametric tests.

Results: Thirty-four patients (13.9%) had diabetes, representing a more aged cohort with a higher prevalence of comorbidities than non-diabetic patients. Pre-dissection RLN amplitude and latency did not differ significantly between groups (0.72 mV vs. 0.74 mV, p = 0.87; 2.2 ms vs. 1.9 ms, p = 0.11, respectively). After dissection, diabetic patients showed a significantly greater reduction in RLN amplitude (median 0.58 mV vs. 0.77 mV in non-diabetics, p = 0.01), whereas latency remained comparable (p = 0.92). Within-group analysis confirmed a significant amplitude drop among diabetics (p = 0.008) but not among non-diabetics. Temporary unilateral vocal fold palsy occurred in three patients (one diabetic, two non-diabetic), all resolving within two weeks.

Conclusions: Diabetic patients demonstrated greater RLN amplitude reduction during thyroidectomy, suggesting impaired neural resilience to surgical manipulation. However, these electrophysiological changes had minimal short-term clinical impact when nerve integrity was preserved. Awareness of this subclinical vulnerability may guide intraoperative decision-making within the context of neuroprotective strategies that involve IONM during thyroidectomy.