Thoracic aortic dissection may present in several ways and requires a high index of suspicion,1,2 as any delay in diagnosis may prove disastrous. It is rare for dissection to present with pulmonary symptoms. Samuel and Goldburgh3 reported four cases who were thought to have pneumonia antemortem but who were found at necropsy to have aortic dissection. We describe a patient who presented with clinical and radiological features suggestive of infective exacerbation of chronic bronchitis or pulmonary embolism but who was found to have aortic dissection.
A 73 year old man with a history of chronic bronchitis secondary to heavy smoking, ischaemic heart disease, and hypertension was admitted to hospital with a two day history of left sided pleuritic chest pain. Two weeks earlier, he developed breathlessness on exertion, fever, and a productive cough not responding to antibiotics. On admission he complained of severe left sided lateral chest pain, aggravated by cough and deep breathing.
On examination he was pale and dyspnoeic with a sinus bradycardia of 55 beats/min, blood pressure of 143/73 mm Hg, and tenderness at the left lower chest wall but no other abnormality. An ECG showed old anterolateral ischaemia and sinus bradycardia, and chest x ray showed cardiomegaly, prominence of the aortic knuckle, unfolding of the descending aorta, and linear atelectasis at the right lung base with pleural thickening. Haematological and biochemical profiles were normal apart from mild renal impairment (plasma creatinine 185 μmol/l). Arterial blood gases showed only hypoxaemia (Po2 8.41 kPa,) so a ventilation/perfusion scan (V/Q scan) was requested; this was done 48 hours after admission but was negative. The patient was treated with oxygen, heparin, simple analgesics, and antibiotics. He remained haemodynamically stable over the following 24 hours and made some improvement but his lateral chest pain was increasing, requiring stronger analgesics. Spiral computed tomography of the chest was then carried out to exclude pulmonary embolism. This showed no evidence of pulmonary embolism but revealed a false lumen in the ascending aorta suggestive of aortic dissection (fig 1). Angiography confirmed dissection of the ascending aorta with minor narrowing of the circumflex artery suggestive of a muscular band.
Spiral computed tomography of the chest following angiography.
At operation diffuse pericardial adhesions with moderate left ventricle hypertrophy but normal right ventricle were noted. There was dissection of the ascending aorta one centimetre proximal to the origin of the innominate artery, extending anteriorly and to the left but stopping short of the supra-sinus ridge. The aortic valve was not involved. There was thrombus in the false lumen. The dissection was grafted but the patient deteriorated shortly after surgery because of intrathorcic bleeding and cardiac tamponade. On second exploration, the bleeding was controlled but he died six days later. Necropsy showed hypoxic encephalopathy with oedematous lungs but an intact anastomatic line at the graft site.
The term dissecting aortic aneurysm, which was coined by Laennec in 1819, has been replaced by aortic dissection, as it is rare for a dissecting aorta to become aneurysmal.4 Historically, aortic dissection was usually a postmortem diagnosis, and this was attributed to the infrequency of the condition, universal lack of clinical suspicion, limitation of investigations, and the absence of a characteristic syndrome.5 Despite the availability of modern radiography, including magnetic resonance imaging, many cases are still misdiagnosed owing to the varied clinical presentation of the condition.
In more than 90% of cases, aortic dissection usually presents with sudden severe chest pain, which may be described as tearing or ripping in nature.4 The pain is often maximal at the time of the onset, and later subsides or remains constant. It is usually very severe; nonetheless many patients may sustain painless aortic dissection.4
Our patient presented unusually with left lateral pleuritic chest pain. Kouchoukos and Dougenis ascribed left pleuritic chest pain to thoracic aortic dissection,6 although this usually occurs with descending rather than ascending dissection. The presence of hypoxaemia with a history of chronic bronchitis made infection or pulmonary embolism more likely than aortic dissection and this was further supported by the chest x ray findings. Jagannath et al studied the x ray findings of 36 patients with angiographically proven aortic dissection,7 and found that widening of the mediastanium and aortic knuckle were the only reliable radiographic features predicting aortic dissection. None of these features was present in our patient.
Patients with aortic dissection usually present with haemodynmic disturbance culminating in circulatory collapse, or with hypertension (which is also a risk factor for aortic dissection) in half to two thirds of cases, especially with distal dissection. The presence of hypotension attributed to rupture of the dissection commonly occurs with proximal dissection.1,4 Compromise of the blood flow to one of the subclavian arteries may lead to variation in blood pressure between the right and left arms.1 This is an invaluable physical sign, which may be missed or ignored. Pulse deficits or absences occur in 50% of patients with proximal aortic dissection, the brachiocephalic artery being more commonly affected.1 This was not detected in our patient although the dissection occurred proximal to the origin of the common brachiocephelic artery.
Although our patient remained hemodynamically stable preoperatively, the poor operative outcome was consistent with a recent finding of Okita et al,8 who reported poor operative outcome in patients over 70 years of age compared with those under 70, especially when operated on as an emergency, and in those with preoperative vital organ damage.
Finally our patient was thought to have pulmonary embolism from his clinical presentation, but a V/Q scan was negative. A case of aortic dissection reported by Stollberger et al had a decrease in perfusion on V/Q scan.9
This case highlights the need to consider aortic dissection as a cause of atypical chest pain. This condition may simulate other pulmonary diseases such as pneumonia or pulmonary embolism.
The principle approach to the diagnosis of acute aortic syndromes is radiographic, though biochemical assays are an area of active research. The electrocardiogram (ECG) can be helpful in detecting rare complications such as coronary involvement; however, it is frequently abnormal and is not specific for acute aortic syndromes [24, 25].
The goals of aortic imaging are fourfold, namely to (1) establish the diagnosis, (2) localize the intimal tear, (3) determine the extent of dissection, and (4) assess for indicators of pending emergency . The modalities available for definitive imaging include computerized tomography angiography (CTA), magnetic resonance imaging (MRI), transesophageal echocardiography (TEE), and, less commonly, aortography. Among the IRAD cohort, CTA was performed in 93% of cases, TEE in 59%, MRI in 31%, and aortography in 24% . Often times more than one study is necessary to confirm and fully characterize the diagnosis; thus the mean number of imaging studies other than plain film performed per patient is 1.8 .
Transthoracic echocardiography (TTE) provides a rapid tool for assessing cardiac complications of type A dissection such as wall motion abnormalities indicative of coronary artery involvement, pericardial tamponade, and aortic insufficiency . TTE provides minimal to no imaging of the distal ascending, transverse, and descending aorta and is typically used in conjunction with TEE or another imaging modality. TEE can be rapidly obtained in many circumstances and has the advantages of excellent visualization of the distal ascending, transverse, and descending aorta, ability to differentiate between true and false lumens, and identification of intramural hemorrhage and penetrating atherosclerotic ulcer. TEE is limited by its invasive nature requiring sedation and a contraindication in significant esophageal disease, such as varices . When combined with TTE, TEE achieves a sensitivity of 99% and specificity of 89%  for aortic dissection.
CT imaging with intravenous contrast can be rapidly performed in most Emergency Departments and has a sensitivity of at least 95% and specificity between 85% and 100% [9, 29, 30]. The disadvantages of CT are the contrast load in patients who already have or may develop acute kidney injury as a result of the underlying process and limited ability to identify the site of intimal tear [9, 31].
MRI has perhaps the highest sensitivity and specificity, approaching 100% for each [27, 29, 32]. MRI has the additional advantages of not requiring intravenous contrast injection for vessel visualization and not exposing the patient to ionizing radiation . Its use is constrained by the factors that limit MRI use in other settings, which are lack of availability, length of time for image acquisition in potentially unstable patients, patient discomfort/anxiety, and contraindication in patients with metallic implants (e.g., most pacemakers and implantable cardioverter-defibrillators).
Given that TEE, CT, and MRI all have favorable test characteristics, the choice of imaging modality largely depends on local availability and time. CT is nearly universally available in the developed world, is non-invasive, and is rapid, and, therefore, is the preferred modality in many institutions . However, some institutions practice routine TEE in patients with proximal dissections. The American Heart Association (AHA) recommends TEE as the first imaging study in unstable patients to prevent patient transport out of the acute care setting .
Biochemical markers of aortic dissection remain largely investigational. Candidate markers include lactate dehydrogenase, d-dimer, white blood cell count, c-reactive protein, fibrinogen, and circulating smooth muscle myosin heavy chain. Currently, there is no clinically available biochemical assay specific for aortic dissection .
The patient described here had CT angiography performed on presentation to the referring hospital, which accurately identified the type B aortic dissection and branch vessel involvement. Transesophageal echocardiography would have been a reasonable initial study as well given the need for close hemodynamic management and the elevated creatinine.