Trans-Oesophageal Migration of Pulled-Out Locking Screw (Zero-Profile Implant System) and its Retrieval Using Suction Catheter: A Technical Note
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.344619
Source of Support: None, Conflict of Interest: None
Keywords: ACDF, pulled-out locking screw, trans-esophageal migration, zero-profile implant
To enhance stabilization, many surgeons prefer to add an anterior cervical plate in cervical fusion procedure., However, the addition of an anterior cervical plate is not without complications. In the early postoperative period, about 2–67% of patients may complain of some sort of dysphagia, and about 3–21% may develop chronic dysphagia-related symptoms. Additionally, the screw-plate interface may lead to cases of migrating screws and subsequent soft tissue damage, along with the higher incidence of adjacent-level degenerations.,,,,,,,,, To prevent these potential complications, a radiolucent implant for stand-alone anterior interbody fusion procedures (Zero-P; Synthes GmbH Switzerland, Oberdorf, Switzerland) of the cervical spine was developed.,, In this technical note, we report a case of dysphagia caused by screw backout from a stand-alone anterior interbody fusion device that subsequently caused esophageal perforation and simultaneously describes an innovative technique of screw retrieval intraoperatively.
A 32-year-old female presented to the outpatient clinic with insidious onset difficulty in walking along with loss of manual dexterity which had worsened over the past 6 months. Neurological examination was suggestive of spastic quadriparesis and sensory abnormalities consistent with the typical clinical syndrome of spondylotic myelopathy. MRI cervical spine showed C5-C6 intervertebral disc herniation with thecal sac compression and local signal changes on T2-weighted MRI images [Figure 1]a. She underwent C5-C6 anterior cervical discectomy and fusion (ACDF) with a variable angle zero-profile anterior cervical interbody fusion (ACIF) device in an uneventful surgery. Postoperative cervical spine X-ray showed well-placed ACIF device and screws [Figure 1]b. She was discharged home on the third postoperative day with mild symptomatic improvement in her preoperative complaints. In the follow-up period, her symptoms resolved completely. Cervical spine X-rays at 3 months were unremarkable [Figure 1]c and follow-up MRI 6 months after the surgery showed a well-decompressed thecal sac with CSF spaces noticeable all around the cord [Figure 1]d.
Nine months following the surgery, she complained of mild discomfort around the operative site in the neck. X-ray cervical spine obtained at this point of time was suggestive of implant loosening and failure of fusion across the instrumented disc space [Figure 1]e. One of the screws seemed to be backing out and endplate erosion along with local cervical kyphosis was noted. Accordingly, a revision surgery was recommended, but the patient refused to undergo further treatment.
However, she returned to the outpatient 6 months later, with increasing local discomfort and mild dysphagia which was more noticeable with liquids. X-Ray of the cervical spine [Figure 1]f and CT cervical spine [Figure 2]a showed further implant loosening, progression of kyphosis, and the screw which had now pulled out completely and now lying in the prevertebral space. Barium swallow showed focal contrast opacification of a hypopharyngeal diverticulum posteriorly at the C5-C6 level. However, no contrast leakage was noted and the barium bolus was seen to descend normally through the hypopharynx, esophagus, and gastroesophageal junction [Figure 2]b and [Figure 2]c.
The patient was taken up for revision surgery after indirect laryngoscopy to assess vocal cord function. Through the previous incision, a standard anterolateral approach was performed to expose the prevertebral space. Dense adhesions were encountered at the level of prevertebral fascia. Implanted zero PEEK cage was found to be loosened and covered with thick granulation tissue which was removed. However, the loosened screw could not be located visually in the initial attempts, though it could be felt by probing the prevertebral space with the index finger. A C6 corpectomy was performed to remove the partially fused PEEK implant. Intraoperative X-rays showed that the screw migrated curiously in the cranial direction as the surgical field was probed further manually [Figure 2]d. After several futile attempts at retrieving the screw, we decided to first fuse the corpectomy level and then resume the search for the screw. Accordingly, an autologous anterior iliac crest was harvested, interposed in the corpectomy space, and buttressed with the anterior cervical plate.
The search for the screw was then resumed. Repeat fluoroscopy after the manipulation showed that the screw had migrated upwards and could be seen at the level of C3 vertebral body [Figure 2]d. At that point, we realized that the probable cause of the unrestricted upward migration of the screw was its possible intraluminal relocation within the esophagus. We then inserted a blunt-tipped suction catheter and applied intermittent negative pressure, further facilitating the ascent of the screw in the oropharynx which was confirmed by X-ray [Figure 2]e. The screw was then finally retrieved using Magill's forceps assisted by video laryngoscopy [Figure 2]f and [Figure 3]a. On further exploration of the wound, a rent of approximately 0.5 cm size was detected in the posterior wall of the hypopharynx at the level of an implant. Esophageal rent repair was done extramucosally with 5-0 PDS and buttressed with sternocleidomastoid rotation flap by our otorhinolaryngologist colleague. Since there was an esophageal perforation detected, the anterior cervical plate was immediately explanted to enable healing of the esophageal rent. The patient was then carefully turned prone and fused aided by C5-6 lateral mass and C7 pedicle screw and rod fixation [Figure 3]b.
Postoperatively, she was then kept on total parenteral nutrition for 2 weeks. She did develop a low-output pharyngo-cutaneous fistula, but it healed gradually with conservative management. She was discharged home on a Philadelphia hard cervical collar when the fistula output was reduced to a trickle. At 1 year of follow-up, she was completely asymptomatic.
Since its initial introduction, ACDF has undergone many technical modifications., One of the major modifications was the introduction of anterior cervical plating., However, dysphagia in the postoperative period was a major concern. To overcome this disadvantage, a new zero-profile device, which combined an anterior plate with a cage (Zero-P, Synthes GmbH, Switzerland) was introduced. Studies have shown it to be a safe and effective method, with similar clinical outcomes as the anterior cervical plating, and a lower risk of postoperative dysphagia.,, Additionally, it is usually associated with shorter operative time, less blood loss, and greater simplicity in use. Very few instrumentation-related complications concerning zero-P have been reported in the literature.,,,, Only three cases of pulled-out locking screws have been mentioned in the literature, as also seen in our case.,,
Screw pull-out strength depends on various factors such as basic screw designs, insertion techniques, bone quality of the individual, bony fusion present or absent, use of orthosis, and so on.,, In our case, the cause of screw loosening and further back out cannot be precisely determined but is probably related to the failure of fusion across the disc space. Nonfused endplates may have led to increased translational motion and instability, abnormally stressing the implant which manifested as a screw pull-out. The initial direction of screw migration was directly impinging upon the posterior pharyngeal wall and could have burrowed into it slowly. Local inflammation and fibrosis would have sealed the perforation and prevented any serious and life-threatening complications such as mediastinitis. A subtle clue to the involvement of the esophageal wall by the displaced implant was the formation of diverticulum in the proximity of the loosened screw. The surgical significance of this finding must be carefully noted. Esophageal diverticula following cervical fusion surgery are frequently associated with implant failure or malposition resulting in chronic irritation of the esophagus., Intraluminal migration of the dislodged screw could be attributed to the intraoperative manipulations. It is instructive to note that extreme mobility of the screw in a vertical direction, observable on serial intraoperative X-rays, was indicative of its intraluminal location.
Therefore, while searching for a screw or implant in a case of extruded implanted instrumentation, intraoperative imaging in the form of C-arm or O-arm is essential.
Once it is ascertained that the implant is inside the esophageal lumen, it can be usually retrieved safely using an upper gastrointestinal endoscopy or flexible bronchoscopy in the same sitting. However, such options need additional specialized equipment and expertise which may not be immediately available. Negative oropharyngeal suctioning performed by the anesthetist and aided by the surgeon can be a safe, efficient, and convenient option in this perplexing scenario. After the retrieval of dislodged screws or implants, it is mandatory to have a meticulous search for any rent or perforations in the esophagus, especially the wall., Most importantly, all ventral metallic implants must be removed to provide a biological surface against which the perforation can heal.
Perforation of the aerodigestive tract by a displaced cervical spine implant is an uncommon complication of instrumented anterior cervical spine fusion. Manipulation during surgery may cause the implants embedded into the wall of the aerodigestive tract to become dislodged and then migrate into the pharyngeal lumen, where they can be easily retrieved using flexible endoscopy. However, in resource-constrained situations or when intraoperative endoscopy is not available, a trial of screw retrieval using a suction catheter may be attempted.
Declaration of patient consent
The authors certify that they have obtained all appropriate patient consent forms. In the form, the patient (s) has/have given his/her/their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3]