Spontaneous bilateral hemothorax in neurofibromatosis type 1 due to internal thoracic artery aneurysm: Case report

Highlights • In patients with NF1, the risk of spontaneous bleeding due to the possibility of aneurysmal formation should be considered.• The differential diagnosis of neurofibromatosis should be investigated in cases of spontaneous hemothorax.• In most cases of spontaneous hemothorax associated with NF1, intercostal artery aneurysm is the most common injury.

We present a case of bilateral hemothorax caused by the rupture of the right internal thoracic artery aneurysm in a patient with NF1.
This work has been reported in line with the SCARE criteria [5].

Presentation of a case
A 43-year-old man, with no past medical history but a family history of neurofibromatosis, was in his usual state of good health until he presented with sudden respiratory distress. He was admitted to the local health service, where his respiratory status deteriorated and endotracheal intubation was needed. Due to a high neck circumference, intubation was difficult and the patient presented continuous hypoxia until successful intubation was finally made after multiple attempts.
He was sent to our hospital after intubation. We admitted the patient hemodynamically stable, sedated, with no need of vasopressors. Physical examination revealed dullness to percussion, decreased breath sounds in both lung fields, multiple cutaneous nodules, café-au-lait spots and Lisch nodules. Thus, we diagnosed type 1 NF1 by clinical criteria and confirmed afterward with the biopsy of one of the skin lesions. On admission, the hemoglobin level was 11.0 mg/dL and then decreased to 7,0 mg/dL in 12 h. Chest computed tomography (CT) with contrast revealed large bilateral hemothorax. Then, bilateral chest drainage was performed and approximately 600 mL of blood was evacuated on the left side, and 800 mL on the right side. Therefore, the patient received isotonic crystalloid and blood transfusion.   He was then submitted to arteriography, which evidenced a right internal thoracic artery aneurysm that was the probable source of the bleeding. However, there was no active bleeding or contrast leak. The aneurysm was selectively catheterized and embolized with coil ( Figs. 1 and 2), followed by a video-assisted thoracoscopy surgery (VATS) to drain the hemothorax, expand the lungs and review the pleural cavity. VATS showed retained and fibrous clot, suggesting chronic bleeding.
Despite therapeutic success in the first 48 h after the intervention, because of the hypoxia during initial intubation, the patient developed cerebral edema that culminated in brain death.

Discussion
NF1 is a multisystem disorder, and there are thoracic manifestations related to neurofibromas in mediastinum, chest wall, lateral thoracic meningocele and vascular structures. Moreover, other manifestations include dysplastic abnormalities of bones and the nervous system [6] and manifested vascular involvement as aneurysm, affecting approximately 3% of the patients [2,3]. Vascular lesions associated do NF1 include stenosis, occlusion, aneurysm, pseudoaneurysm, and rupture [7]. The mechanisms differ according to the size of the vessel: In large vessels, aneurysms are formed by the direct invasion of the vascular wall by tumors tissue, which presses the vasa vasorum and causes ischemia and weakness. Smaller vessels suffer dysplasia, that initiates stenosis of the wall, weakens and causes friability [2].
In this case report, the classical cutaneous findings suggested the diagnosis hypothesis of NF1 for the patient. When there is a vasculopathy, signs and symptoms vary depending on the size and on the location of the lesion [2]. This patient had an acute onset of dyspnea and the increase of the cervical volume, due to the hematoma. However, initial symptoms could not be better investigated because the patient was admitted at the hospital already unconscious.
In this case, we found bilateral hemothorax in chest CT, but no suggestive etiology for the bleeding. Arteriography, on the other hand, showed an aneurysm in the right internal thoracic artery. The cause of the bilateral hemothorax was considered to be chronic bleeding of this aneurysm, explained by its degeneration and fragility, due to NF1. The bleeding dissected to both pleural cavities caused the bilateral hemothorax and impaired ventilation chronically. Chest CT was carefully interpreted looking for neurofibromas which could also be a cause of the hemothorax, as suggests previous case report, six but there were no other findings.
Endovascular embolization is indicated if the patient is hemodynamic stable. In unstable patients, it is recommended aggressive treatment, thoracotomy with surgical ligation [2,4]. VATS is believed to be the best available modality for the management of clotted hemothorax [4]. The treatment of massive hemothorax is by thoracotomy or VATS, to identify the cause and solve it. The surgical vessel reconstruction is limited because the arteries of patients with NF1 are fragile [10]. Previous published cases suggest that coil embolization has better results [2,9,10].
As the patient of this case was stable, the medical team opted to perform an endovascular embolization of the aneurysm and then use VATS to drain the pleural cavities and expand the lungs. VATS, then, showed retained and fibrous clot, suggesting chronic bleeding. The aneurysm embolization was effective in preventing new bleeding.
However, even with successful endovascular intervention and pulmonary expansion, the patient had a hypoxic-ischemic cerebral injury and died. It must be remembered that these are critical patients, and hemostasis does not always mean the resolution of the problem, once there can be many associated complications. The disease mortality is 36% and postoperative mortality is 33% in patients with hemothorax associated with NF [2].

Conclusion
The differential diagnosis of neurofibromatosis should be advanced in cases of spontaneous hemothorax. In patients diagnosed with neurofibromatosis, the risk of spontaneous bleeding due to the possibility of aneurysmal formation should be considered.

Sources of funding
No sponsor involvement in the study.

Ethical approval
The research was approved by the Research Ethics Committee of the Federal University of Vale do Sao Francisco protocol number 3649509 with the consent of University Hospital of the same university.

Consent
Written informed consent was not obtained from the patient. The head of our medical team has taken responsibility that exhaustive attempts have been made to contact the family and that the paper has been sufficiently anonymised not to cause harm to the patient or their family. A copy of a signed document stating this is availble for review by the Editor-in-Chief of this journal on request.

Author contribution
Antonio

Registration of research studies
The Plataforma Brasil is a database for recording human research. Our search has the following record: CAEE: 20876719.5.0000.5196.

Guarantor
Paulo Fernandes Saad. Head of residency in cardiovascular surgery.

Provenance and peer review
Not commissioned, externally peer-reviewed.

Declaration of Competing Interest
There is no conflict of interest.