Incidental finding of large pneumothorax on Cardiac MR scan

Pneumathoraces are potentially life threatening causes of chest pain with a rapid diagnosis critical to improving the chances of a positive outcome. They classically present with acute pleuritic chest pain and dyspnoea with the diagnosis confirmed on chest X ray. Treatment may be conservative or involve needle decompression depending on individual clinical impact. Although very rarely identified on MR imaging, supervising staff and reporting doctors should be aware of their key features. This case serves to identify not only the abnormal lung parenchymal features but also the striking compressional effect of a pneumothorax on the right ventricle during both systole and diastole. Indeed, we believe these are the first published images of a pneumothorax identified on CMR.

A 26-year-old presented for follow up Cardiac Magnetic Resonance (CMR) of his thoracic aorta. He had undergone an end-to-end anastomosis repair of a thoracic aortic co-arctation aged 3. His last CMR scan 7 years previously had demonstrated normal appearances to the repair. The patient had not reported any symptoms at a recent clinic appointment and did not complain of any symptoms when he attended for imaging. He underwent successful CMR without incident and subsequently returned home.

The scan was reported 24 h later by the cardiology fellow who suspected a pneumothorax and after confirming their suspicion with the on-call Radiologist immediately phoned the patient to take a further history. This revealed that the patient had in fact felt slightly short of breath for the preceding few weeks associated with an uncomfortable sensation in his back on coughing and inspiration. The patient was asked to return to the emergency department for further clinical assessment where a chest X ray revealed a 3 cm apical pneumothorax. Despite the striking compressional effects of the pneumothorax on the patient’s right ventricle (RV), the patient was clinically well with normal observations and a clinical decision was made to manage the pneumothorax conservatively with close follow up. After 3 weeks and two interval chest X rays, the pneumothorax was found to have resolved.

This case highlights several important messages. Firstly, whilst pneumothoraces are only rarely diagnosed on CMR, they are an important, potentially life threatening condition that should not be overlooked. Anyone reporting CMR scans should therefore be aware of the key imaging features (as demonstrated in Fig. 1). Moreover, there is an argument that recognition should be made by supervising doctors at the time of scanning allowing suspected cases to undergo appropriate additional imaging and immediate clinical assessment. Furthermore, this case demonstrates that when a pneumothorax is detected by CMR this modality is actually an excellent technique with which to assess its haemodynamic compression upon the right ventricle in diastole (Additional files 1, 2 and 3).

Finally, this case illustrates the importance of paying close attention to extra-cardiac structures whilst reporting CMR scans, with close radiology support for reporting cardiologists [1]. Although examination of extra-cardiac structures is widely adopted by most imaging cardiologists it is not a current CMR training requirement [2] with most imaging centres providing a co-reporting service consisting of a cardiologist and radiologist. This acts to ensure no major extra-cardiac abnormalities are overlooked. Major non-cardiac findings are however only uncommonly discovered on CMR. Chan et al. [3] studied 1534 clinical CMR scans and reported the prevalence of non-cardiac pathology as being under 10% with major findings reported as 3% and major new findings less than 0.5%. Dewey et al. [4] studied 108 patients undergoing clinical CMR scans and similarly found non-cardiac findings prevalent in 7% with only 2% significant non-cardiac findings [4]. McKenna et al. [1] reported a higher prevalence of non-cardiac pathology among a group of 107 volunteers for CMR research, demonstrating potentially significant pathology in 17%. Strikingly not one of the patients in these studies (n = 1749) was found to have a pneumothorax. Indeed, we believe these are the first published images of a pneumothorax identified on CMR and certainly the first to illustrate compromised RV filling.