Answers

Short answers

1. Aortic dissection—dissection of the arch of the aorta and the carotids. This led to the unusual presentation with headache.

2. Predisposing factors are hypertension (72% of cases), atherosclerosis (31%), and previous cardiac surgery (18%). Marfan’s syndrome, bicuspid aortic valve, or previous aortic surgery are more likely causes in younger patients.

3. Laboratory investigations: creatine kinase, troponin T/I, myoglobin, D-dimers, haematocrit. Diagnostic tests: electrocardiography, chest radiography, transoesophageal echocardiography, and computed tomography or magnetic resonance imaging of the chest.

4. The patient is initially managed by obtaining good intravenous access and cross matching at least six to eight units of blood. Intravenous β blockers should then be started, and if the ascending aorta is affected the cardiothoracic surgeons should be contacted as a matter of urgency.

5. Clinical predictors—Marfan’s syndrome, female sex, history of atherosclerosis or a previous aortic aneurysm, and the presence of pleural effusion. Anatomical predictors—a patent false lumen, a maximal aortic diameter of >4 cm at presentation, and an intramural haematoma with localised dissection or ulcer-like projections on the computed tomography scan.

Long answers
1.Diagnosis
Aortic dissection. This occurs when the integrity of the intimal layer is breached, thereby allowing blood at high pressure to penetrate the media (figure). Mortality is about 33% within the first 24 hours and 50% within 48 hours if left untreated.¹ Better diagnostic and management strategies have improved the survival of patients with hypertension, which has led to an increase in the rate and prevalence of aortic disease.² Between 1987 and 2002 the incidence of aortic disease rose by 52% in men and 28% in women to reach 16.3 per 100 000 per year and 9.1 per 100 000 per year, respectively.³

View larger version (32K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Cardiac magnetic resonance imaging (three dimensional, contrast enhanced magnetic resonance angiography aortogram), GE 14X 1.5 Tesla, gadolinium contrast (0.2 mg/kg intravenously).The large arrow (left) is the brachiocephalic artery. The small arrow (right) is the intimal dissection flap in the descending aorta (type B aortic dissection). Courtesy of Vincent L Sorrell, University of Arizona College of Medicine, USA

 
Aortic dissection can present with a variety of symptoms, the most common being chest pain, stroke, or congestive cardiac failure.⁴ Typically, the pain starts suddenly and is intense. The site of the pain depends on the location of the tear. If the tear is above the aortic valve the pain is felt anteriorly, but if it is distal to the left subclavian vein it is felt posteriorly.⁵

Classic physical findings like aortic regurgitation and pulse deficit are noted in only 31.6% and 15.1% of patients, respectively.⁶ The prevalence of a difference in blood pressure between the two arms ranges from 12% to 18.4% for a systolic difference 20 mm Hg and 13% to 33.7% for a diastolic difference 10 mm Hg.⁷ If the dissection affects the ascending aorta, the left subclavian ostium may become occluded. This will cause the blood pressure in the left arm to decrease and result in a difference in blood pressure between the arms. The table shows details of the signs and symptoms and their causes.

View this table: [in this window] [in a new window]   Signs and symptoms of aortic dissection

 
Headache or neck pain has been a presenting symptom of spontaneous proximal aortic dissection in a few case reports.^{8 9 10 11} When the dissection extends to the carotid or vertebral arteries, the presenting symptom can be headache, and ischaemic stroke can occur. Headache as the initial manifestation of aortic dissection may result from distension of the vessel wall or ischaemia of the pericarotid plexus.⁹ It can also be caused by transient malperfusion of the carotid arteries. Opisthotonos and neck stiffness may occur as a reaction to severe pain. These symptoms may mimic signs of subarachnoid haemorrhage but without classic symptoms of aortic syndrome.¹¹

2. Causes of aortic dissection
The International Registry of Aortic Dissection (IRAD) study showed that predisposing factors were hypertension (72% of cases), atherosclerosis (31%), and previous cardiac surgery (18%). Further analysis showed that younger patients were less likely to have hypertension (34%) or atherosclerosis (1%) but were more likely to have Marfan’s syndrome, bicuspid aortic valve, or previous aortic surgery (box).¹²

Causes of aortic dissection Common causes Hypertension Atherosclerosis    Connective tissue disorder Marfan’s syndrome Ehlers-Danlos syndrome    Hereditary vascular diseases Biscuspid aortic valve Coarctation of aorta    Inflammation of the vasculature Temporal arteritis or Takayasu arteritis Behçet’s disease Syphilis    Trauma Motor vehicle crash Fall from a height    Iatrogenic factors Catheter or instrument intervention Valvular surgery or aortic surgery Side clamping or cross clamping Aortotomy Graft anastomosis Patch aortoplasty Cannulation site or aortic wall fragility    Other causes Pregnancy Turner’s syndrome Smoking Dyslipidaemia Use of cocaine

3a. First line investigations
Several first line investigations are available. If laboratory tests, chest radiography, and electrocardiography are inconclusive, computed tomography or transoesophageal echocardiography are the definitive tests.

• Laboratory tests
  

Some laboratory tests are performed largely to narrow the differential diagnosis towards aortic dissection. For example cardiac enzymes, creatinine kinase for myocardial infarction, and D-dimer for pulmonary embolism. Other laboratory tests like crossmatching, measurement of haemoglobin, and haematocrit investigate the possibility of dissection.

• Chest radiography
  

Aortic dissection should be suspected when chest radiography shows mediastinal widening. However chest radiography is neither specific nor sensitive for the diagnosis of aortic dissection.

A study of 216 patients suspected of having acute aortic syndromes over a six year period reported a sensitivity of 67% for overt aortic dissection and of 63% for intramural haemorrhage or penetrating ulcer.¹³ The sensitivity was 47% when the aortic dissection was proximal and 77% when distal.¹³

• Electrocardiography
  

All patients with suspected aortic dissection should undergo electrocardiography to help exclude acute myocardial infarction.^{4 14} It is also important to remember that the dissecting aortic membrane may extend into a coronary ostium (usually that of the right coronary artery) causing acute myocardial ischaemia.

Results of electrocardiography are normal in a third of patients with coronary involvement and most of these patients have non-specific ST-T segment changes.¹⁴ However in some patients with coronary involvement electrocardiography changes look the same as for acute ST elevation myocardial infarction, which could result in detrimental administration of thrombolytic treatment.

About 20% of patients with type A dissection have evidence of acute ischaemia or acute myocardial infarction on electrocardiography.¹⁴ If dissection is suspected in a patient with signs of myocardial infarction on electrocardiography, diagnostic imaging must be performed to rule out dissection before thrombolytic treatment is considered.[]

• Computed tomography
  

Computed tomography is the definitive test for diagnosing aortic dissection. Multislice computed tomography can help identify the intimal flap, the extent of the dissection, branch vessel involvement, the size of the aorta, the patency of the false lumen, and pericardial effusions including those involving the proximal coronary arteries.¹⁵

Identification of the intimal flap separating the true and false lumens is essential for the diagnosis of aortic dissection. Intramural haematoma appears as a crescent shaped, high attenuation signal within the wall of the aorta on non-contrast computed tomography. It can also appear as a localised thickening of the aortic wall with internal displacement of intimal calcifications.^{15 16}

The sensitivity of computed tomography for the diagnosis of aortic dissection ranges from 83% to 100%, and the specificity ranges from 87% to 100%.^{15 16} Other findings suggestive of aortic dissection are outer wall calcification, intraluminal thrombus, and eccentric flap calcification. If computed tomography isn’t helpful, transoesophageal echocardiography and magnetic resonance imaging, which both have a high sensitivity, can be used. A comparative study found that computed spiral tomography, multiplanar transoesophageal echocardiography, and magnetic resonance imaging had a specificity of 100%, 94%, and 94%, respectively, in aortic dissection.¹⁷ When assessing involvement of the aortic arch vessel, these tests had a sensitivity of 93%, 60%, and 67%, respectively, and specificity of 97%, 85%, and 88%.

• Transoesophageal echocardiography
  

Transoesophageal echocardiography overcomes many of the limitations of transthoracic echocardiography because of the proximity of the oesophagus to the aorta. This is widely available, relatively safe, and easy to perform at the bedside. The sensitivity of transoesophageal echocardiography for aortic dissection has been reported to be as high as 98%, and specificity ranges from 63% to 96%.¹⁸ However, patients may need to be sedated, which may have unwanted haemodynamic effects in unstable patients.

It is used to visualise not only the intimal flap within the aortic lumen with high spatial resolution but also flow within the false lumen.

3b. Second line investigation

• Magnetic resonance imaging
  

The sensitivity and specificity of magnetic resonance imaging for diagnosing aortic dissection range from 95% to 100%.¹⁹ The advantages of this test are its non-invasiveness, high resolution, three dimensional reconstruction, use of less toxic gadolinium as contrast agent, and absence of ionising radiation. The disadvantages are inconvenience, lack of availability on an emergency basis, and concerns about patient monitoring and prolonged scanning time, especially in unstable patients. Several recent metallic implants limit the use of magnetic resonance imaging—contraindications include pacemakers, implantable cardiac defibrillators and the Starr-Edwards’s prosthetic valve.²⁰

4. Management
Patients suspected of having aortic dissection require large bore intravenous access for fluid and drug administration. Adequate pain control with intravenous morphine is necessary to allay anxiety and to reduce sympathetic stimulation, which may in turn raise the blood pressure.

The systolic blood pressure should be reduced to a target of 110 mm Hg or the lowest that is tolerated without signs of hypoperfusion.²¹ Intravenous β blockers reduce the aortic shear stress and are therefore the first line drugs for reducing blood pressure in aortic dissection.⁴ Propranolol (0.05-0.15 mg/kg every four to six hours) or labetalol can be used initially. Labetatol can be given as a bolus or infusion (20 mg bolus followed by 20-80 mg every 10 minutes to a total dose of 300 mg or an infusion at 0.5-2 mg/min).

If blood pressure remains high after adequate β blockade at its maximum dose or to a heart rate <60 beats/min, a vasodilator can be added to the treatment regimen. Direct vasodilators should not be used alone, because they can cause an increase in sympathetic drive that will increase the force of left ventricular ejection and risk further dissection.⁴ Sodium nitroprusside (initial dose of 0.25 µg/kg/min and then titrated to target blood pressure) is the preferred vasodilator in combination with a β blocker. Angiotensin converting enzyme inhibitors are an alternative to nitroprusside.²¹

Patients with aortic dissection of the ascending thoracic aorta (DeBakey types I and II, or Stanford type A) have a high risk for early death, because the dissected ascending aorta has a propensity to rupture into the pericardial space. Consequently, such patients are managed with early surgical repair of the ascending thoracic aorta, as decades of data have validated the advantage of surgery over medical treatment in this group.

Patients with aortic dissections of the descending aorta but not the ascending aorta (DeBakey type III, Stanford type B) have a relatively low risk of aortic rupture and, thus, have a significantly lower early mortality. Moreover, surgery to repair the descending thoracic aorta in this setting is associated with substantial mortality and morbidity, including paraplegia. Because surgery confers no survival advantage over medical treatment for uncomplicated type III aortic dissection, medical therapy is the treatment of choice.²² Early surgery to repair the descending thoracic or abdominal aorta is reserved for patients in whom vascular complications of dissection, such as rupture or malperfusion syndromes, may arise.

In a series of patients with repaired type I aortic dissection, 78% of patients were alive at five years and 66% at 10 years compared with about 93% and 80% for an age and sex matched control population.²³ Data from the international registry of acute aortic dissection show that 77% of patients with medically managed type III aortic dissection were alive at three years.²⁴ Survival was higher in a prospective study of medically managed acute type III aortic dissection in Sweden—82% at five years and 69% at 10 years—and only slightly below that of the age and sex matched general population, and most of the patients died of causes unrelated to their dissections.²⁵

The successful endovascular stent grafting repair of both acute complicated and chronic stable type III aortic dissections was reported in 1999.^{26 27} Since then, the technique has become more widely used in place of open surgery for the treatment of the vascular complications of acute dissection. Although randomised trials have not been conducted, the available data suggest that early mortality is considerably lower for stent grafting than for open repair in patients with vascular complications of type III aortic dissection.²⁸ Stent grafting is not without risks. In a recent meta-analysis of 39 studies of 609 patients with distal dissection treated with stent grafts, major complications occurred in 11%, with stroke or paraplegia in 3%.²⁸ Moreover, complications were seen in 21% of patients with acute rather than chronic dissection. A recently completed randomised European trial (investigation of stent grafts in patients with type B aortic dissection) compared the two year outcome of endovascular stenting versus medical treatment for uncomplicated subacute and chronic (two weeks to 12 months) type III dissection.²⁹ Its preliminary results showed that, surprisingly, stent grafting had no survival benefit after the first year of the two year follow-up. Indeed, before considering routine use of stent grafts in patients with uncomplicated type III aortic dissection, it should be recognised that most patients do relatively well with medical treatment alone. Hence, identifying those patients with distal dissection who may benefit from prophylactic stent grafting requires improved methods of risk stratification.

5. Clinical and anatomical predictors of increased risk of death after aortic dissection
Clinical predictors of increased risk of death include Marfan’s syndrome, female sex, history of atherosclerosis or a previous aortic aneurysm, and the presence of pleural effusion. Several anatomical features of the dissected aorta are predictive of increased risk of death: a patent false lumen,²³ a maximal aortic diameter of >4 cm at presentation,^{23 25} and presence of an intramural haematoma with localised dissection or ulcer-like projections on the computed tomography scan. Patients with pure intramural haematoma and no evident communication with the aortic lumen are at low risk.²⁵

One report concluded that a large false lumen was the strongest predictor of distal aortic expansion in patients with chronic type I aortic dissection.³⁰ However, these investigators identified a large false lumen by measuring the cross sectional area of the false lumen divided by the area of the whole aorta (both lumens) at that level, to provide a ratio. They concluded that a ratio of 0.7 is predictive of progressive aortic enlargement. Although their findings are interesting, this method of sizing the false lumen is cumbersome in clinical practice.

Cite this as: BMJ 2008;337:a1844