Esophageal injury related to ventral spine surgery is fortunately rare. The largest study assessing the incidence of this complication was performed by Newhouse et al. The authors surveyed the membership of the Cervical Spine Research Society about their experiences with esophageal perforation following ventral cervical spine surgery. The response rate was 56%. They identified 22 cases of perforation among an estimated 10,000 cases of ventral spine surgery for an incidence of 0.25% (1). Gaudinez et al. (2) found 44 cases of esophageal perforation in a select population of 2,946 patients (1.49%) with cervical fracture and a spinal cord injury. Orlando et al. (3) reported on 1,075 cases of ventral cervical surgery performed in their clinic and noted five cases of esophageal or piriformis sinus perforation for an incidence of 0.4%. Patel et al. (4) reported 3 cases of esophageal perforation in 3,052 operations (0.1%) over 34 years (see Table 111.1).

A number of risk factors have been implicated in esophageal injury. Newhouse et al. reported that 6 of the 22 cases they identified involved intraoperative injury by sharp or motorized instruments. Another 4 of the 16 cases identified after surgery were felt to be related to hardware use (1). Patel et al. (4) noted that ventral instrumentation had been used in all of their cases with a perforation. Several case reports have proposed an erosive mechanism of the esophagus over either prominent hardware or bone graft (5,6). Mendoza-Lattes et al. (7) described ischemic changes in the esophagus due to retractor placement during ventral cervical surgery. Although their focus was dysphagia, rather than esophageal injury, persistent ischemia may produce tissue necrosis.

Cervical trauma also seems to be a significant risk factor. Comparing large series, the incidence in the review by Gaudinez et al., who looked specifically at trauma patients, is substantially higher (1.49%) than reported in other studies (0.1% to 0.4%) (1, 2, 3 and 4). The authors noted several cases of direct injury to the esophagus as part of the original trauma but attributed 77% (34/44) of their cases to the surgical procedure rather than the trauma (2). Newhouse et al. (1) noted that the most common indication for surgery among patients with a reported perforation was trauma and postulated that fracture was an independent risk factor for esophageal injury. Nerot et al. (8) reported a case of a cervical extension fracture treated without surgery, which developed fever and respiratory distress on postinjury day 4 and was subsequently diagnosed with esophageal perforation, which they suspected was caused by compression of the esophagus against the ventral vertebral body.

There has been a suggestion that the esophagus may be most vulnerable to injury in the region from C5 to C7. Newhouse et al. found that 50% (11/22) of the esophageal perforations reported in their study occurred at C5-C7. They were unable to define whether this was related to the frequency of surgery at those levels or increased risks associated with surgery in this region (1). Orlando et al. reported 60% (3/5) of the injuries in their series occurred at these levels. Gaudinez et al. did not specify the levels of surgery in their study but reported that 88% (39/44) of their patients had C4-C6 level injuries, which presumably means surgery at these levels as well. Patel et al. noted that all three of their cases occurred at C5-C6 and postulated that the dorsal esophageal mucosa is especially thin at this level.

Esophageal injury carries a significant morbidity. Gaudinez et al. reported an average length of hospital stay of
253 days in their patients with esophageal perforation. Their patients suffered numerous additional medical complications, including atelectasis, pneumonia, airway obstruction, malnutrition, and pulmonary embolism. Complications requiring additional surgery included esophageal strictures (59%), cervical osteomyelitis (50%), prevertebral or retropharyngeal abscess (31%), and fistula formation (36%). Septicemia developed in 16% of patients and mediastinitis in 11% (2). In the cervical spine research society (CSRS) series, 88% (14/16) of cases identified after surgery required reexploration. The remaining two cases required feeding tubes and prolonged intravenous antibiotics (1). While some case reports describe essentially asymptomatic perforation of the alimentary tract and screw migration (6,9), more recent reports describe a clinical course similar to that reported by Gaudinez and Newhouse, with prolonged antibiotics, wound exploration, and feeding tube placement (4).

Historically, the mortality rate from perforation of the cervical esophagus has ranged from 15% to 50% (1,10). The spine surgery literature, however, does not support so high a rate. Newhouse et al. (1) reported a 4.5% (1/22) mortality rate from sepsis. Gaudinez et al. (2) reported the same mortality rate (2/44), without specifying the cause. Several smaller series of esophageal perforations related to spine surgery contain too few patients to report a mortality rate (3, 4, 5 and 6,9,11). Among the 13 cases described in these reports, no mortalities are reported. Certainly, these injuries result in significant morbidity, and this analysis is limited by reporting bias and small numbers. Nonetheless, the historically quoted mortality rate of 50% for these injuries (1,3,4) appears to be higher than the current literature supports (see Table 111.1).