Dysphagia, hematoma, vocal cord paralysis, dural injury, Horner syndrome, and esophageal perforation are some of the postoperative complications of anterior cervical spine surgery. Of these, esophageal perforation is a rare but fatal complication, with mortality rates between 6% and 34%. Because of its rarity, with an incidence ranging from 0.2% to 3.4% [1,2], most spine surgeons have limited experience in the treatment of esophageal perforation. To date, only a few case studies have been reported in the literature. Herein, we report the causes, diagnosis, and management of esophageal perforation, depending on the time of diagnosis.

This retrospective study was approved by the Institutional Review Board of the Uijeongbu St. Mary's Hospital (e-IRB UC18RESI0153) and informed consent was waived. In total, seven patients treated by seven different spine surgeons were analyzed. The demographic characteristics of the patients are detailed in Table 1, and all patient data were reviewed retrospectively. In all patients, primary cervical spine lesion, primary cervical surgery type, risk factors, predicted causes of esophageal injury, duration from first symptom to surgery, period from surgery to diagnosis, reason for diagnostic delay, diagnostic method, coinfection, identification of organism, treatment method, and treatment results were examined. Based on the time of esophageal perforation, the patients were categorized into the following groups: the intraoperative group (diagnosed intraoperatively), perioperative group (diagnosed within 30 days postoperatively), and delayed group (diagnosed >30 days postoperatively).

Esophageal perforation may not be immediately identified in anterior cervical spine surgery; hence, a later possibility must be considered. The esophagus is located in the superficial layer of the deep cervical fascia, running on the back of the organ and surrounded by smooth muscle [3]. Although dysphagia is the most common symptom of esophageal perforation, other symptoms, such as pharyngeal pain, odynophagia, unexplained aspiration, chocking, fever, localized neck tenderness, induration, hemoptysis, and subcutaneous emphysema can also occur [4,5]. Food residue, such as rice leaking from the surgical incision, is a pathognomonic finding. In the present study, dysphagia was the main symptom in all patients, and two patients were diagnosed by food residue drainage performed 3 days postoperatively. The risk factors for esophageal perforation were surgery (reoperation of the cervical spine, use of high-speed burr, use of sharp retractor, implant migration, and prolonged operative time) and patientrelated factors (diabetes mellitus or tumor). In this study, the cause of esophageal perforation, classified by time of onset, was infectious spondylitis in the intraoperative group, revision surgery (one case of spondylitis and one case of metal failure) in the perioperative group, and chronic maceration of metal after traumatic fracture surgery in the delayed group.

Secondary infections are another main symptom of esophageal perforation, but delayed infections because of esophageal perforation are not characterized by symptoms. Broad-spectrum antibiotics, which are sensitive to both gram-positive and -negative organisms, should be administered, even before the identification of bacteria [6]. In this study, methicillin-resistance Streptococcus epidermidis was identified in one of the two patients in the intraoperative group. The patient eventually died of sepsis caused by uncontrolled infection. Streptococcus epidermidis and S. viridans were identified in one of two patients in the perioperative group. The patients in the delayed group had no associated infections.

Plain neck radiographs may be helpful in visualizing indirect signs of perforation, such as the presence of prevertebral air, subcutaneous emphysema, widening of the retropharyngo-esophageal space, and migration of cervical implants. Barium esophagogram (contrast swallow study) may help confirm the diagnosis and locate the perforation, showing extravasation of fluid and/or air fluid collection in deep neck spaces [7]. Although esophagoscopy can provide direct visualization of the perforation, it may miss perforation hidden in a mucosal fold, leading to high false-negative rates (10%–32%) [8]. In our patients, esophagography was the most common confirmative diagnostic method (six cases), but it produced a falsenegative result in one case. Endoscopic examination was the confirmative diagnostic method in three cases, with no false-negative findings. One case was confirmatively diagnosed by direct visualization through secondary exploration.

Management of esophageal perforation may vary based on the time of diagnosis, size and shape of injury, accompanying surgical site infection, and the patient’s general condition. Multidisciplinary cooperation between otolaryngologists, gastroenterologists, spine surgeons, and cardiothoracic surgeons is essential. The mortality rate is 20% if treated within 24 hours, but is increased to 50% if delayed >24 hours [9,10]. Small perforation, a well-contained leak with no sign of sepsis, is only possible with conventional treatment [11,12]. Conventional therapy requires prohibition of oral administration for at least 1 week, administration of broad-spectrum antibiotics, administration of prokinetic drugs (cisapride and metoclopramide), and feeding via nasogastric tube [11]. However, 20%–25% of patients develop abscesses because of treatment failure even with conventional treatment, resulting in a mortality rate of 18% [10,13].

Primary closure of perforation and prevention of infection is the gold standard of treatment if a surgical procedure is required [13]. If primary closure is impossible, sternocleidomastoid muscle flap, pectoralis major muscle flap, or a longus coli muscle flap can be reinforced after a double-layer or imbricating suture. Supportive treatment, in addition to surgical treatment, is also important. Jejunostomy is recommended when the nasogastric tube is maintained for at least two weeks [11]. In the intraoperative group, anterior plate and screw removal and posterior instrumentation were performed. Esophageal perforation was treated with reconstruction using the omentum flap in one patient and primary repair in one patient. Esophageal perforation was treated conservatively after implant removal surgery in all patients in the perioperative group. Anterior plate and screw removal was performed in all three patients in the delayed group. Esophageal perforation was treated by primary closure in two patients who were diagnosed 3 months postoperatively. Esophageal perforation was treated by reconstruction using the sternocleidomastoid muscle flap in a patient who was diagnosed 32 months postoperatively. Drainage was performed in all cases, and the mean duration of drainage was 5.6 days (range, 2–12 days). One patient developed sepsis postoperatively, followed by disseminated intravascular coagulation, and died. The remaining six patients were cured without serious neurologic sequelae.

The retrospective, multicenter design of the present study is a fundamental limitation. Our conclusions may not be strongly appealing to other spine surgeons and practitioners. We understand that the drawback of our study is inevitable because esophageal perforation is a rare complication occurring after anterior cervical spine surgery.

The main cause of intraoperative esophageal perforation was esophageal adhesions because of infectious spondylitis. However, perioperative and delayed esophageal perforations were caused by chronic irritation of metal failure. Anterior plate and screw removal is necessary, and posterior instrumentation and fusion may be considered in these patients, depending on the fusion status. Repair method for esophageal perforation should be determined by considering the severity of esophageal perforation.