Fever of Undetermined Origin: Cervical Diskitis-Osteomyelitis Associated With Esophageal Self-Expanding Metal Stent

Abstract

Conflicts of interest The authors disclose no conflicts. © 2013 by the AGA Institute 1542-3565/$36.00 http://dx.doi.org/10.1016/j.cgh.2013.05.039 Aof an esophageal self-expanding metal stent (SEMS). She was diagnosed with squamous cell carcinoma of the upper esophagus 6 months previously at another facility. She underwent placement of the partially covered SEMS at that time. She subsequently underwent chemotherapy and radiation therapy, with follow-up positron emission tomography–computed tomography (CT) showing no evidence of residual disease. About 2 weeks after completion of therapy, she started experiencing intolerance to liquids in the form of coughing episodes, followed by onset of fevers. Work-up for fever at the outside facility was unrevealing. Endoscopic removal of the SEMS was unsuccessful, and she was referred to our facility. Upon review of systems, she also reported upper back pain for the past few weeks. A barium esophagogram did not reveal any leak, fistula, obstruction, or perforation. A CT scan of the thorax suggested erosion of the posterior aspect of the SEMS into the C7 and T1 vertebral bodies (Figure A, sagittal CT view). Endoscopic views of the SEMS were unremarkable. The SEMS was removed under general anesthesia using a double-channel therapeutic gastroscope and 2 rat-tooth forceps, with no evidence of perforation noted after stent removal (Figure B). A barium esophagogram obtained after stent removal showed no evidence of a leak or fistula. Subsequent magnetic resonance imaging of the C-spine and T-spine showed enhancement of the C7, T1, T2, and T3 vertebral bodies, suggestive of diskitis-osteomyelitis (Figure C, sagittal T2-weighted magnetic resonance image). Blood cultures were negative. The patient’s fevers abated with appropriate antibiotic therapy. At the last follow-up evaluation 3 months later, she remained asymptomatic without any back pain or fevers. Esophageal SEMS have become an integral part of the management of malignant dysphagia. The patient’s initial symptoms of cough and intolerance to oral intake were suspicious for a tracheoesophageal fistula. However, barium esophagogram before and after stent removal failed to show the same. Her symptoms were likely caused by cervical spine involvement, esophageal dysmotility, and/or reflux. Cases of spondylodiscitis and osteomyelitis associated with esophageal stents, both plastic stents and SEMS, for benign indications as well as malignant indications have been reported previously. Although most cases have occurred in association with overt stent-induced perforation, there was no evidence of perforation in the current case. It is possible that a small perforation may have sealed by the time of fluoroscopic and endoscopic study at our institution. Given the widespread use of SEMS for palliation of dysphagia, clinicians need to be aware of the above described delayed complication associated with their use.