MON-553 Nerve Injury after Ethanol Ablation of Thyroid Cancer Metastases to the Neck

Background Thyroid cancer often recurs in the neck. Well-differentiated thyroid cancer metastasizes to the neck in about 40%, and recurs after surgery in about 20%-30%. Surgery is the treatment of choice, with biochemical remission in 21-66%, but some patients are poor surgical candidates or refuse surgery, and permanent nerve injury and hypoparathyroidism are substantial complications. The 2018 National Comprehensive Cancer Network (NCCN) guidelines state that in select patients with unresectable, non-radioiodine-avid, and progressive disease, local therapies such as ultrasound guided ethanol ablation (EA) can be considered. Recent systematic reviews have confirmed the efficacy and safety of EA, and show that this minimally invasive therapy compares favorably to surgical resection in non-randomized studies. Objectives To determine the rate, type, and outcomes of nerve injury after EA for neck metastases from thyroid cancer. Methods IRB approval was obtained. Retrospective review of medical records of all EA performed from March 2004 through August 2018. Data collected included sex, age, histologic type of thyroid cancer, date of all EA sessions, and number of tumor sites injected per EA session. Nerve injury was defined as a neurologic deficit with immediate onset after ethanol injection that lasted beyond the effect of local anesthetic. Diagnosis was made based on symptoms and functional deficit as described during follow up phone calls or visits. Confirmatory laryngoscopy was not performed. Results During the 14 year period 104 patients received EA; 64 females with mean age 60 (range 23-90) and 40 males with mean age 54 (range 24-85). A total of 298 sites were treated (lymph nodes and local extra-nodal recurrence) in 169 ethanol injection sessions (mean 1.8, range 1-7). Histologic diagnosis was papillary in 89.4% (classic 87, follicular variant 4, tall cell variant 2), medullary 6 (5.8%), and follicular 4.8% (follicular 1, Hurthle cell variant 4). Recurrent laryngeal nerve (RLN) deficit occurred in 2, superior laryngeal nerve (SLN) deficit in 1 and sympathetic deficit in 1. Nerve injury occurred in 4/104 patients (3.8%), 4/169 EA sessions (2.4%), and 4/298 EA sites (1.3%). All EA with nerve injury were in level VI, including a very deep injection associated with the sympathetic chain injury. The dysphonia from two RLN injuries recovered fully at 1 week and at 3 months. One SLN injury had no hoarseness but could not project a loud voice while singing with his church choir and recovered in 1 month. The one patient with ptosis recovered almost completely, did not notice any residual problems, but physical examination demonstrated a subtle difference in palpebral elevation 1 year after EA. Conclusion Nerve injuries after EA are infrequent and most resolve completely within 3 months.