Tracheocarotid artery fistula in a patient who had tracheostomy successfully treated with a saphenous vein graft
- Katherine Elizabeth Dahill and
- Ged Dempsey
- Critical Care, University Hospital Aintree, Liverpool, UK
- Correspondence to Dr Katherine Elizabeth Dahill; katherine.dahill@nhs.net
Abstract
Tracheoarterial fistula is a complication of tracheostomy with a high associated mortality. A 25-year-old male patient with Duchenne’s muscular dystrophy underwent a percutaneous tracheostomy using the single tapered dilator (Blue Rhino) technique to facilitate weaning from mechanical ventilation. Nine weeks after the procedure, he developed significant upper airway bleeding, leading to haemodynamic instability. A CT angiogram of the neck and thorax did not reveal a source of the bleeding. The patient was subsequently transferred to the operating theatre where a 1 cm defect in the right common carotid artery was found and repaired with a graft from the left short saphenous vein. Clinicians who undertake tracheostomy formation should be aware of the possibility of tracheoarterial defect and may wish to discuss it at tracheostomy formation. It should be considered early in the event of a significant bleed. This case identifies deep tissue infection and misplacement of the tracheostomy tube as major contributing factors to fistula formation.
Background
Tracheoarterial fistula (TAF) is a rare complication of tracheostomy with a reported incidence of <1.0%1 and 0.35% in the intensive care unit (ICU) in question.2 However, due to the potential for rapid deterioration and high associated mortality,3 it is important to recognise and manage appropriately. Although the most commonly implicated artery is the innominate, this and several other case reports illustrate that the common carotid artery can also become fistulous with the trachea.4 5 This is important to bear in mind when approaching a patient who has a tracheostomy with an ongoing major haemorrhage. Bronchoscopy and CT angiography are the most useful diagnostic tools, but as demonstrated here, they are not always reliable.
Case presentation
A 25-year-old man with a history of Duchenne’s muscular dystrophy presented with respiratory failure due to pneumonia and progressive muscular weakness. Prior to admission, he required overnight non-invasive ventilation. He was a wheelchair user and hoist transferred.
This patient was admitted to the ICU for invasive ventilation. A percutaneous tracheostomy, using the single tapered dilator, was sited early in anticipation of a prolonged wean from respiratory support. The tracheostomy was inserted with bronchoscopic control. An 8 mm dual-lumen cuffed Portex tube was inserted with no immediate complications. The patient was subsequently transferred to the hospital weaning unit for further management.
Nine weeks after tracheostomy insertion, the patient developed significant bleeding from the nose and mouth. Despite ongoing resuscitation, he developed systemic hypotension and ventricular dysrhythmias requiring direct current cardioversion.
Initially, the nasal cavities were packed as this was felt to be the most likely source of bleeding. The airway remained patent and free of blood. The tracheostomy tube was left in situ throughout with the cuff inflated.
On arrival at the critical care unit ongoing bleeding from the site of the tracheostomy became evident.
Investigations
An arterial blood gas taken at the time of initial haemorrhage demonstrated a profound lactic acidosis (serum lactate 12 mmol/L (0.5–2.2 mmol/L)) and a pH of 6.7 (7.360–7.420).
CT angiography of the neck and superior mediastinum after stabilisation revealed no obvious source of haemorrhage. It did, however, reveal that the tracheostomy tube was misplaced and appeared to be deviated to the right hand side, causing it to lie in close proximity to the brachiocephalic trunk (figure 1). There was also pooling of fluid below the vocal cords to the level of the tracheostomy, representing blood.
CT angiogram of the neck taken while bleeding ongoing. Blue arrow - tracheostomy tube.
Differential diagnosis
The most likely differential diagnoses were epistaxis and tracheoinnominate artery fistula.
As the majority of blood was being lost via the nose and packing of the nasal cavity appeared to slow the bleeding, massive epistaxis was felt to be more likely.
As there was no haemorrhage into the airway initially, it was felt that a TAF was a less likely differential.
Treatment
The patient was subsequently transferred to the operating theatre with input from cardiothoracic, ear, nose and throat and vascular surgeons. Resuscitation continued during surgery with concomitant vasopressor therapy.
Following removal of the tracheostomy tube, the bleeding point was noted to be arising from within the trachea. Further exploration revealed erosion of the right lateral wall of the trachea into the right common carotid artery, where a 1 cm defect was found.
The carotid was cross-clamped and a patch repair of the defect was undertaken using the short saphenous vein.
Tissue samples taken intraoperatively grew Streptococcus milleri, and the patient was treated with the appropriate antibiotics (amoxicillin and metronidazole) for a total of 6 weeks.
Outcome and follow-up
Twelve days after the repair, the patient returned to theatre for surgical tracheostomy formation. The trachea was seen to be healed, with no evidence of a defect.
The patient subsequently exhibited evidence of left-sided neglect, and a CT of the brain revealed a right middle cerebral artery infarct.
Four months after the initial bleed, the patient remains ventilator-dependent in hospital.
Discussion
The primary mechanism of injury leading to TAF formation is pressure necrosis from an overinflated or malpositioned cuff. Contributing factors include mucosal trauma, low tracheal incision and radiotherapy.6 Patients with neurological problems are more likely to develop TAF,7 due to a variety of factors, including higher probability of long-term tracheostomies. One paper reports that skeletal deformity in patients with Duchenne’s muscular dystrophy may contribute to the misalignment of tracheotomy tubes,8 a point that should be considered in the aetiology of this case.
This event was likely multifactorial, with the primary cause being misplacement of the tracheostomy tube to the right as demonstrated in the CT images. The majority of TAFs are reported to occur 3 days to 6 weeks post procedure,6 with later bleeding more often attributable to causes such as granulation tissue or malignancy.6 In this patient, pre-existing respiratory muscle weakness resulted in a prolonged weaning time from mechanical ventilation. An extended period of time with the cuff of the tracheostomy tube up and causing pressure damage to the adjacent structures eventually lead to fistula formation.
The majority of cases discussed in the literature are tracheoinnominate artery fistulas, with cases of tracheocarotid fistula (TCF) being rare but not unheard of. The reason for this is anatomical, with the innominate artery lying directly anterior to the trachea in the majority of cases, around the level of the ninth costal cartilage.9 The right common carotid arises from a bifurcation of the innominate artery and is contained in the carotid sheath lateral to the trachea. Unusually, the deviation of the tube here caused lateral necrosis of the right tracheal wall adjacent to the right carotid artery.
S. milleri are a group of commensal oral organisms which have been known to become pathogenic, forming abscesses and causing local extension of infection in the head and neck.10 11 While the presence of positive tissue culture is most likely to represent colonisation, it cannot be ruled out that infection contributed to the tissue damage and subsequent fistula formation. Continuous pressure, in addition to infection-driven inflammation, results in friable tissues, which in combination with a low tracheostomy site may result in TCF.12 Patients with long-term tracheostomies should be continuously monitored for clinical signs of infection, to avoid catastrophic sequelae such as this.
The majority of reported cases are managed surgically, if the patient is stable enough for the theatre. The management of TAF, following initial resuscitation, involves immediate bronchoscopy and overinflation of the tracheostomy cuff. If bleeding into the airway is evident on bronchoscopy, then translaryngeal intubation and digital compression of the artery is advised.6 In this case, bronchoscopy was not performed as the tracheostomy site was clear of blood and the airway was patent. As the blood appeared to be coming from the oropharynx, the priority was the patient’s cardiovascular status as he rapidly became peri-arrest due to hypovolaemia. The cuff was not overinflated for the same reason. When it became clear that the bleeding was from the stoma, the patient was transferred to theatre for surgical exploration. The tracheostomy tube placement in relation to the fistula ensured that the patient’s airway was protected throughout.
Options for repair of TAF include patch closure of the artery in question or replacement with a venous graft. One reported case of TCF was successfully repaired with endovascular stent implantation,13 which may be preferable in certain cases. This would, however, require a clearly identified bleeding point on the imaging.
The bleeding point not being visible on CT may be due to valve formation causing an intermittent bleed. Had the bleeding point been visible, it is likely that the volume would have been such that patient would have exsanguinated prior to scanning. Depending on the situation, immediate surgical exploration may be preferable to waiting for imaging in a patient with an ongoing massive bleed.
Learning points
-
There should be a high index of suspicion of an arterial fistula in a patient who had tracheostomy and who develops haemorrhage more than 3 weeks after insertion. The priority is to maintain oxygenation and resuscitate adequately while planning definitive management.
-
A multidisciplinary approach is required for management, with experienced anaesthetics input throughout.
-
Patients with prolonged cannulation and prolonged cuff-up weaning time are at increased risk of tracheoarterial fistula.
-
When planning further tracheostomy formation, consider healing time of the initial defect alongside the long-term consequences of prolonged intubation.
Footnotes
-
Contributors I, KD, was the first author of this article. I planned and wrote the report. The second draft was written with the help of GD, who advised on how to phrase the report and what made the case stand out. I got the images from the radiology department with the help of a consultant radiologist.
-
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
-
Competing interests None declared.
-
Provenance and peer review Not commissioned; externally peer-reviewed.
- © BMJ Publishing Group Limited 2021. No commercial re-use. See rights and permissions. Published by BMJ.
References
Use of this content is subject to our disclaimer