Endoscopic endonasal resection of the odontoid process: clinical outcomes in 34 adults

OBJECTIVE Treatment of odontoid disease from a ventral corridor has consisted of a transoral approach. More recently, the endoscopic endonasal approach (EEA) has been used to access odontoid pathology. METHODS A retrospective review was conducted of patients who underwent an EEA for odontoid pathology from 2004 to 2013. During our analysis, the mean follow-up duration was 42.6 months (range 1-80 months). Patient outcomes, complications, and postoperative swallowing function were assessed either by clinic visit or phone contact. RESULTS Thirty-four patients underwent an EEA for symptomatic odontoid pathology. The most common pathology treated was basilar invagination (n = 17). Other pathologies included odontoid fractures, os odontoideum, and metastatic carcinoma. The mean patient age was 71.5 years. Thirty-one patients underwent a posterior fusion. All 34 patients experienced stability or improvement in symptoms and all had successful radiographic decompression. The overall complication rate was 76%. Nearly all of these complications were transient (86%) and the overall complication rate excluding mild transient dysphagia was only 44%. Twenty-one patients (62%) suffered from transient postoperative dysphagia: 15 cases were mild, transient subjective dysphagia (6 of whom had documented preoperative dysphagia), whereas 6 other patients required tube feedings for decreased oral intake, malnutrition, and dysphagia in the perioperative setting (5 of these patients had documented preoperative dysphagia). Sixteen patients had documented preoperative dysphagia and 6 of these had lower cranial nerve dysfunction. Postoperatively, 6 (37.5%) of 16 patients with preoperative dysphagia and 4 (67%) of 6 with lower cranial nerve dysfunction had significant dysphagia/respiratory complications. Eighteen patients had no documented preoperative dysphagia and only 2 had significant postoperative dysphagia/respiratory complications (11%). The rates of these complications in patients without preoperative dysphagia were lower than in those with any preoperative dysphagia (p = 0.07) and especially those with preexisting lower cranial neuropathies (p = 0.007). Dysphagia was also significantly more common in patients who underwent occipitocervical fixation (19/26, 73%) than in patients who underwent cervical fusion alone or no fusion (2/8, 25%; p = 0.02). All patients with perioperative dysphagia had improved at follow-up and all patients were tolerating oral diets. No patient suffered from velopalatal insufficiency. Two patients had intraoperative CSF leaks. One of these patients underwent a negative exploratory surgery for a questionable postoperative CSF leak. One patient developed infection in the resection bed requiring debridement and antibiotics. One patient died 8 days following surgery from an unknown cause. The 90-day perioperative mortality rate was 2.9%. CONCLUSIONS A completely EEA can be performed for compressive odontoid disease in all cases of neoplastic, degenerative, or invaginative atlantoaxial disease with satisfactory outcomes and low morbidity. Transient perioperative dysphagia and respiratory complications are common, usually as an exacerbation and reflection of underlying disease or occipitocervical fusion rather than the EEA, emphasizing the importance of avoiding transoral surgery.