Discussion

The Capp's triad, described in 1934, refers to the commonly observed clinical and radiological findings in patients with retropharyngeal haematomas: subcutaneous bruising over the anterior neck and chest, anterior displacement of the trachea, and tracheal or oesophageal compression resulting in airway or swallowing problems.^{Reference Capps11} Although Capps initially described ‘mediastinal and subcutaneous haemorrhage’, his descriptions mirror today's terminology of ‘retropharyngeal haematomas’. Capps proposed spontaneous bleeding from one of multiple asymptomatic parathyroid adenomas as the cause,^{Reference Capps11} also described by Hellier and McCombe in 1997 in a patient with a mildly raised serum calcium level.^{Reference Hellier and McCombe12} Whilst our case demonstrated very mildly elevated PTH levels initially, the calcium levels remained normal and there was no biochemical evidence of parathyroid adenoma. Thus, the authors suggest that the mildly raised PTH may have arisen because the blood temporarily irritated one or more of the parathyroid glands, later self-resolving.

Anatomically, Penning et al. suggested that many cases of retropharyngeal haemorrhage may be secondary to bleeding from vessels overlying the anterior longitudinal ligament.^{Reference Penning13} Other authors proposed the thyrocervical trunk or its branch, the inferior thyroid artery, as a rarer source.^{Reference Van Valde, Sars, Olsman and Van de Hoeven14,Reference Calofero, Miller and Greenberg15} The authors propose that the combination of indomethacin affecting platelet aggregation, alongside the patient coughing (at the time of her symptom onset), may have led to this retropharyngeal haematoma; the coughing mechanism could have caused shearing forces as a result of attachments of the nearby fascial layers.

Consensus for likely bleeding sources has not been achieved, but Pfeiffer and Ridder described utilising a transcervical approach to: achieve endotracheal intubation, identify the bleeding defect, and enable insertion of a haemostatic agent and drain to surgically manage a retropharyngeal haematoma.^{Reference Pfeiffer and Ridder7} Other authors have performed CT angiograms to assess the bleeding location and evaluate its suitability for surgical intervention or embolisation.^{Reference Pfeiffer and Ridder7,Reference Betten and Jaquint10,Reference Van Valde, Sars, Olsman and Van de Hoeven14}

Fortunately, surgical evacuation or transoral aspiration is rarely necessary, and would likely only be undertaken in cases of rapidly expanding or non-resolving haematomas.^{Reference Lin and Sinclair5} In patients using oral anti-coagulants and anti-platelet agents, reversal agents such as vitamin K and fresh frozen plasma should be considered, amongst other newer reversal therapies, to minimise bleeding.^{Reference Betten and Jaquint10,Reference Bloom, Haegen and Keefe16} However, reversal agents were not employed in this case, as indomethacin was the only potentially contributory medication and the patient was stable for 5 days prior to admission.

The literature describes varied management strategies for retropharyngeal haematoma with and without airway compromise. In this case, there was no evidence of airway compromise; in such instances, the majority of authors advocate conservative management, in the form of observation and antibiotic coverage, as was followed.^{Reference Duvillard, Ballester and Romanet3,Reference Lin and Sinclair5} Although steroids and antibiotics are frequently used, their benefit is unclear.^{Reference Bloom, Haegen and Keefe16} With rarer impending airway obstruction, some have advocated primary tracheostomy to avoid the potential, but unreported, risk of disrupting the retropharyngeal haematoma through endotracheal intubation.^{Reference Senthuran, Lim and Gunning4,Reference Van Valde, Sars, Olsman and Van de Hoeven14}

Ongoing management can include serial CT imaging. However, the authors felt the clinical stability and resolution of this case with supportive measures did not warrant repeat imaging, and its associated radiation exposure.^{Reference Lin and Sinclair5} In conjunction with the literature, the patient was admitted for observation as, although documented in post-trauma patients, authors have warned of delayed stridor and airway compromise.^{Reference Lazott, Ponzo, Puana, Artz, Ciceri and Culp17} Our case was rare in its atraumatic nature, and no literature exists regarding follow-up duration in such cases. Comparatively, Penning studied 30 patients hospitalised with cervical spine injuries, and found that most retropharyngeal haematomas self-resolved within two weeks of the injury.^{Reference Penning13} Another study described resolution of traumatic retropharyngeal haematoma within four weeks.^{Reference Mitchell and Heniford18} This may suggest an appropriate guidance follow-up period. In our atraumatic case, the patient reported near-complete symptom resolution within two weeks.

This is the first reported atraumatic retropharyngeal haematoma presumed to be secondary to indomethacin use. The most similar cases include a report by Simckes et al., who described a peri-renal haematoma and irreversible renal failure following ketorolac use for pain management in sickle cell disease.^{Reference Simckes, Chen, Osorio, Garola and Woods19} Additionally, Buckley et al. used electron microscopy in rat models to demonstrate reduced platelet aggregation and significantly prolonged bleeding times after just one dose of ketorolac.^{Reference Buckley, Davidson and Das20} Of interest, Armstrong et al. described a mediastinal haematoma in a patient taking ketoprofen, with vigorous stirring of a Christmas pudding described as contributory, due to the presumed shearing forces generated.^{Reference Armstrong, Kirkby-Bott and Johnson21} Ketorolac and ketoprofen are NSAIDs, similar to indomethacin; perhaps, alongside our case, these studies suggest NSAIDs as a potential risk factor for atraumatic haematomas in various anatomical locations.

Using our experience alongside the literature, we would suggest conservative management with observation and close discussion with the haematology department, providing there are no concerns regarding airway compromise or a rapidly enlarging haematoma. With naturally resolving symptomatology, we found serial imaging was not necessary. However, calcium and parathyroid hormone levels should be measured on admission, and any abnormalities must be investigated to exclude spontaneous haemorrhage from a parathyroid adenoma, as has previously been described.