Esophageal diverticula could be either true diverticula (involving all the layers of the esophagus) or false diverticula (involving only the mucosal and submucosal layers only), pulsion diverticula (resulting from increased intraluminal pressure due to esophageal dysmotility leading to esophageal wall herniation in an area of weakness) or traction diverticula (wherein mediastinal inflammation adheres and pulls on the esophageal wall from outside causing a defect or diverticula). Based on the anatomic location, esophageal diverticula are further divided into three types namely [6]; (1) Zenker’s diverticulum at the pharyngoesophageal junction (pulsion or false diverticula occurring in the Killians triangle), (2) Rokitansky diverticulum or mid-esophageal diverticulum which are true diverticula arising either due to traction or pulsion, and (3) epiphrenic diverticulum within 10cm above the gastro-esophageal junction (pulsion or false diverticula that may result from increased lower esophageal sphincter pressure e.g. in achalasia).
Rokitansky or mid-esophageal diverticula were historically considered to be traction diverticula secondary to mediastinal inflammation from granulomatous infections like tuberculosis [7] and sarcoidosis [8] in the mediastinal lymph nodes causing local traction on the esophageal wall. However, recent studies have shown that a majority of mid-esophageal diverticula are pulsion diverticula secondary to motility disorders [9, 10, 11, 12, 13]. In addition, tuberculosis which was one of the primary causes of traction diverticula, is in decline in the Western world. Although most patients with esophageal diverticula are asymptomatic, some present with dysphagia, regurgitation, aspiration pneumonia, weight loss, heart burn, chest pain and vomiting. Long standing inflammation can lead to a full thickness erosion of the esophageal wall leading to esophagobronchial fistulas, mediastinitis [4] and any perforation or leak can lead to direct spread of GI microbes into the prevertebral space potentially leading to spondylodiscitis or spinal epidural abscess [13].
Esophageal diverticulum requires a thorough work-up including an esophagogram (to rule out mechanical obstruction), manometry (to rule out motility disorders), pH manometry (to rule our GERD) and endoscopy (to rule out mucosal lesions like neoplasms and ulcers). The treatment of esophageal diverticula is based on the pathophysiology and natural history of the disease: (a) asymptomatic diverticula do not need a specific treatment, (b) small diverticula may be left in place and not resected, (c) medium-size diverticula may be either treated by diverticulectomy, diverticulopexy, or esophagodiverticulostomy in case of pharyngoesophageal diverticula, (d) resection is probably the ideal therapy for larger diverticula, and (e) a myotomy should always be included to the procedure [2]. Minimally invasive procedures like video-assisted thoracoscopic surgery, laparoscopy and combined video-assisted thoracoscopic surgery and laparoscopy to perform a diverticulectomy and myotomy have been shown to be safe, successful [14] with reduced postoperative morbidity compared with the open thoracotomy approach [1]. Freeman has shown excellent outcomes following the deployment of esophageal stents for esophageal perforation, fistula, or anastomotic leak [15].
Spinal infection has been reported following iatrogenic esophageal injuries resulting from endoscopic management of Zenker’s diverticulum [16, 17], upper GI endoscopic procedures [18], esophagectomy surgery [19], concurrent chemotherapy and radiotherapy [20], stenting of esophageal strictures [21], accidental esophageal injuries while eating fish [22] and chronic esophageal perforation [23]. Esophageal injuries have also been reported following anterior cervical spine surgery leading to infective spondylitis and spinal wound infection (0.3–0.9%) [24, 25, 26].
A 56-year-old gentleman developed pyogenic cervical spondylitis four months after nasogastric feeding for a severe head injury. He was also found to have a Zenker’s diverticulum. The spinal infection healed after a combination of surgical repair of the diverticulum and course of antibiotics [27]. Another 69-year-old lady who presented with scapular pain was diagnosed an esophageal diverticulum with fistula and T2-T3 spondylodisciitis. The patient refused an esophageal surgery and subsequently developed altered sensorium and encephalitis and became ventilator dependent [28].
The ‘‘classic triad’’ of localized back pain, progressive neurologic deficit, and fever in the diagnosis of spinal epidural abscess occurs only in 37% of patients [29], temperature above 37.5°C in 13–68% [30, 31], neurological symptoms in 33% [30, 31], constitutional symptoms (anorexia, nausea, vomiting and weight loss) in 5–50% [31, 32], leukocytosis (> 11500/cumm) in 34–65% and elevated ESR (70 mm/hr) in 72% of patients [30, 31]. However, the time to diagnosis of spondylodisciitis and epidural abscess is an independent risk factor for the development of adverse outcomes (1.4 months versus 2.1 months) [33] and a delay in diagnosis usually varies from 1 to 6 months [30, 34]. Hence a strong index of suspicion is necessary to make an early diagnosis of infective spondylodisciitis and prevent a poor outcome.
The patient discussed in this report was admitted with infrascapular back pain without any neurological symptoms/signs, fever, elevated blood counts or inflammatory markers. However, the CT scan revealed an esophageal diverticulum with an esophago-spinal fistula and presence of contrast in the T2-T3 disc space. An MRI scan with contrast revealed reduced T2-T3 disc height with end plate and marrow oedema, uniform contrast enhancement of T2-T4 vertebral bodies and a small epidural collection suggestive of an infective spondylodisciitis. Although a CT guided percutaneous needle biopsy did not reveal the infective organism, a blood culture grew Streptococcus Pseudoporcinus which is a facultative Gram-positive coccus characterized by a large zone of beta-haemolysis. In the absence of spinal instability and symptomatic cord compression, it was decided to manage the spinal infection with appropriate antibiotics (based on sensitivity studies) and treat the esophageal diverticulum simultaneously. The patient’s medical co-morbidities precluded an esophageal diverticulum repair surgery and hence deployment of an endoluminal self-expanding metallic stent was considered. Fortunately, a combination of esophageal stenting and extended course of antibiotics led to healing of the infective spondylodisciitis and correction of the leaking diverticulum.