Thoracoacromial artery injury after tube thoracostomy for pneumothorax

  1. Shravan Leonard-Murali ,
  2. Adhnan Mohamed ,
  3. Ann Woodward and
  4. Dionne Blyden
  1. Surgery, Henry Ford Hospital, Detroit, Michigan, USA
  1. Correspondence to Dr Shravan Leonard-Murali; smurali1@hfhs.org

Publication history

Accepted:07 Jul 2020
First published:17 Aug 2020
Online issue publication:17 Aug 2020

Case reports

Case reports are not necessarily evidence-based in the same way that the other content on BMJ Best Practice is. They should not be relied on to guide clinical practice. Please check the date of publication.

Abstract

In this case, a patient presented in a delayed fashion after blunt trauma is found to have a large left-sided pneumothorax, and tube thoracostomy is performed. After placement of the apically oriented tube, he developed haemothorax. CT imaging showed an area of questionable extravasation from the left subclavian artery, directly anterior to the thoracostomy tube. His haemothorax was refractory to adequate drainage with a new thoracostomy tube. He ultimately required angiography, coil embolisation and covered stent placement, followed by thoracoscopic evacuation of the haemothorax.

Background

Haemothorax after tube thoracostomy for pneumothorax is very uncommon, with the most often cited aetiologies being intercostal or lung parenchymal injuries. Injury to the subclavian artery or its branches is infrequent. Here we report a case of significant haemothorax after tube thoracostomy for simple pneumothorax following blunt trauma. We were fortunate that our delay in diagnosis did not cost the patient his life, but the situation is a reminder that the uncommon does occasionally happen, and should be considered when common aetiologies are ruled out.

Case presentation

A 30-year-old man presented to the emergency department with persistent chest pain and shortness of breath after being assaulted by multiple individuals 1 day prior. He had no prior medical or surgical history and did not report any symptoms aside from those mentioned. Physical examination was notable for normal vital signs and reduced breath sounds on auscultation of the left haemithorax. Laboratory values (complete blood count, arterial blood gas, serum lactate) were within normal limits, including a haemoglobin value of 136 g/L. Chest X-ray was notable for a large left-sided pneumothorax (figure 1), for which a tube thoracostomy was performed in the fourth intercostal space using a 20 French apically oriented tube (figure 2). The thoracostomy tube did not contain a trocar and was made of pliable, clear, polyvinyl chloride. The procedure was uncomplicated and completed in standard fashion, under local anaesthetic, with a small incision, convincing and controlled pleural entry with Kelly clamp, and introduction of the thoracostomy tube into the pleural space. There were no obvious fractures or other abnormalities. Of note, the thoracostomy tube was abutting the apex of the left haemithorax, enough that it created a buckle in the contour of the tube. After placement, the patient had some relief of his shortness of breath and vital signs remained stable. He was then transferred from the initial emergency department, which was a smaller scale facility, to a tertiary care centre for further management. On arrival, he did not report any new symptoms, aside from some discomfort with the thoracostomy tube. Vital signs remained stable and within normal limits. His thoracostomy tube had drained 200 mL of dark blood to this point.

Figure 1

Initial chest X-ray on presentation to the emergency department demonstrating large left-sided pneumothorax.

Figure 2

Chest X-ray after insertion of first thoracostomy tube, with resolution of pneumothorax but blunting of left costophrenic angle. The area of buckled tube is circled in red.

Investigations

Initial investigations on presentation to the tertiary care centre included basic laboratory tests (complete blood count, arterial blood gas, serum lactate), which were only notable for a decreased haemoglobin from his previous value at the initial emergency department (haemoglobin 103 g/L). Due to the bloody drainage from his thoracostomy tube and the haemoglobin decrease, a chest X-ray was repeated and showed blunting of the left costophrenic angle (figure 3). Contrast-enhanced CT of the chest was then performed, demonstrating a large left haemothorax and an area of radiographic abnormality posterior to the left subclavian artery (figure 4). After discussion with the radiologist, the finding was deemed to most likely be artefact, but extravasation could not be excluded as a possibility. We did not obtain a multiphase (peak arterial, peak venous, late venous) protocol CT scan as the contrast-enhanced images were obtained as part of a broader trauma imaging protocol and the radiologist felt more strongly that the finding was artefact rather than true extravasation.

Figure 3

Progression of costophrenic blunting suggesting haemothorax on chest X-ray after transfer to tertiary care centre.

Figure 4

Contrast-enhanced CT scan demonstrating area of possible extravasation from left subclavian artery branch. Of note, this was not a multiphase protocol CT scan. Red arrow points to extravasation. Thoracostomy tube is posterior to this area and haemothorax at apex of haemithorax is evident.

After performance of a new tube thoracostomy using a 36 French tube, 1.2 L of dark blood was evacuated and the index 20 French thoracostomy tube was removed. The new tube was set to −20 mm Hg suction. The patient was followed with serial haemoglobin values, which did gradually decrease (haemoglobin 79 g/L). Additionally, serial chest X-rays were obtained and notably demonstrated an initial resolution of the haemothorax (figure 5), followed by recurrent pneumothorax and haemothorax preventing lung apposition to the chest wall (figure 6). During this time, the thoracostomy tube continued to drain bloody fluid at a rate exceeding 200 mL/hour, before eventually ceasing to drain due to presumed clot burden. It was decided at this time that further investigation would occur in the operating room, in the form of angiography and thoracoscopy. Angiography did reveal extravasation from the thoracoacromial artery (figure 7). This was treated with coil embolisation and covered stent placement (figure 8). Thoracoscopy after this did not reveal any ongoing bleeding or significant trauma to the apex of the haemithorax.

Figure 5

Chest X-ray after 36 French thoracostomy tube placement with evacuation of haemothorax.

Figure 6

Chest X-ray a few hours after larger bore thoracostomy tube placement with failure of lung apposition to chest wall despite continued drainage and tube being affixed to −20 mm Hg suction.

Figure 7

Diagnostic angiogram demonstrating extravasation inferior to left subclavian artery, most convincingly from the thoracoacromial artery. The area of extravasation is circled in red.

Figure 8

Completion angiogram after coil embolisation and covered stent placement.

Differential diagnosis

The differential diagnosis evolved throughout the patient’s clinical course. On arrival to the tertiary care centre, we presumed that the patient had sustained a bleb rupture from blunt trauma and had been adequately treated with his index tube thoracostomy at the initial emergency department. However, the patient’s haemothorax on subsequent imaging dispelled this notion, and the questionable left subclavian artefact noted on contrast-enhanced CT chest imaging did alert us to the possibility of arterial pseudoaneurysm rupture. Additionally, the large amount of blood evacuated with the 36 French tube thoracostomy signalled some haemorrhagic events, though it was still unclear whether this was a delayed presentation from the blunt traumatic event or it was related to the tube thoracostomy insertion at the initial emergency department. Though it became obvious, this patient required treatment in the operating room, due to the potential for extravasation from the left subclavian artery, we opted to begin with angiography instead of thoracoscopy. This was done in order to adequately treat this possible life-threatening diagnosis in the case it was confirmed.

Treatment

Initial treatment focused on relief of his pneumothorax with tube thoracostomy. This was successful but led to development of haemothorax on the same side. Considering the lack of opacity on his first chest X-ray (which showed his pneumothorax), and blunted costophrenic angle on the subsequent chest X-ray shortly thereafter, it is likely his haemorrhage was incited by the tube thoracostomy placement at the initial emergency department. Also possible is that his blunt traumatic event formed a pseudoaneurysm, which was ruptured by the index thoracostomy tube. We then treated his haemothorax with a larger, more inferiorly placed thoracostomy tube, ensuring it did not abut the apex of the haemithorax. The index thoracostomy tube was removed as well. This confirmed our suspicion of ongoing haemorrhage but did not solve the problem.

Definitive treatment would only be possible in the operating room, but due to the artefact which was concerning for extravasation on the CT images, we elected to perform angiography first, especially considering the initial CT was not done with a multiphase protocol. Extravasation from a thoracoacromial arterial branch was noted and stopped with coil embolisation and placement of a covered stent. Completion angiogram confirmed all extravasation had ceased. Finally, in order to fully evacuate the chest of blood and allow lung expansion, we performed thoracoscopy. Haemothorax of 2.5 L was evacuated, and it was noted that the large-bore thoracostomy tube was well placed but clotted. The chest was examined carefully and no active bleeding was noted, including from the apex of the haemithorax. Two large-bore thoracostomy tubes were placed under direct vision in an apical and basilar orientation, respectively, and the patient was returned to the intensive care unit for monitoring.

Outcome and follow-up

The patient recovered well following surgery, with complete resolution of his symptoms of shortness of breath and chest pain. He did not develop any symptoms in his left upper extremity, and only required aspirin 81 mg /day as prophylaxis for stent patency. His haemoglobin level was stabilised. Thoracostomy tubes were trialled off suction by postoperative day 2 and sequentially removed on postoperative days 3 and 4. He was discharged home with minimal pain and complete resolution of symptoms on postoperative day 5 and was noted to be doing well on clinical follow-up. His stent remains patent and he has not had any recurrence of symptoms.

Discussion

Haemothorax after tube thoracostomy for simple pneumothorax is uncommon, with only other case reports in the literature. In the setting of some other potential causes (such as trauma in this case), other aetiologies should be investigated first. However, if they do not adequately explain the clinical picture, tube-induced injury must be considered. During standard tube thoracostomy placement, there are several steps during which complications can occur. Those most frequently reported involve malposition or compression of structures.1–3 Damage to intercostal vessels, lung parenchyma or the heart is less common.4–7 Even less frequently reported, however, are cases of subclavian vasculature injury after tube thoracostomy.8–10

Case reports of subclavian branch haemorrhage after tube thoracostomy are varied in presentation and onset of haemorrhage, but most cite onset as being delayed.8–10 Theories regarding the mechanism involve suction trauma to small vessels or formation of pseudoaneurysm during initial placement followed by delayed rupture.10 In our patient, the position of the index thoracostomy tube was suspicious, as it was pushed up against the apex of the haemithorax. Despite the fact there were no signs of obvious injury to the area during thoracoscopy, this remains our most highly suspected aetiology for the haemorrhage, as the onset of haemorrhage was immediately after placement. One final theory that we considered but ultimately felt was unlikely that his blunt traumatic event formed a pseudoaneurysm, and the thoracostomy tube ruptured it. However, considering his lack of any fractures, no outward bruising or areas of tenderness, and physically fit nature, we felt his trauma was unlikely to have been significant enough to cause an isolated thoracoacromial arterial pseudoaneurysm. In any case, diagnosis can quickly and effectively be obtained with contrast-enhanced CT, preferably with a multiphase protocol. A multiphase protocol CT scan could have been obtained immediately after the radiologist noted the area of abnormality noted on the initial images, and there is one area for possible improvement in this case.

The final decision to make was regarding treatment of the apparent subclavian artery injury. We considered endovascular, open and thoracoscopic approaches. There have been multiple reports of each approach being successful in experienced hands, with choices being made based on personnel available and the condition of the patient.11 The extravasation appeared to be from a thoracoacromial arterial branch and he was a young, healthy patient without previous vascular pathology, so we felt that an endovascular approach would cause less morbidity and would be a safe starting approach. Fortunately, it was successful, and he avoided the need for further open or prolonged thoracoscopic exploration.

Learning points

  • When performing tube thoracostomy, avoid pushing the tube too far into the chest. This will prevent pressure injury to the apex of the haemithorax.

  • Haemothorax after tube thoracostomy for simple pneumothorax should be further investigated, preferably with high-resolution cross-sectional imaging with a multiphase protocol.

  • If there is copious bloody drainage from thoracostomy tube, and extravasation is noted on imaging from central vessels, angiography should be considered either before or after thoracic exploration.

  • Timing of angiography should be determined on a case-by-case basis and involved multidisciplinary discussions with radiology, interventional radiology or vascular surgery and trauma surgery teams.

Footnotes

  • Contributors All authors were intimately involved in the care of the patient. SL-M wrote the article and accumulated the images. AM assisted with image accumulation. AW and DB provided senior guidance and editing support.

  • 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.

  • Patient consent for publication Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

References

Use of this content is subject to our disclaimer