The primary mechanism of injury is acute deceleration with a variety of forces applied to the descending thoracic aorta near the ligamentum arteriosum. Recently, increased attention has been paid to the “osseous pinch” mechanism. Traumatic rupture had been considered an absolute surgical emergency, with immediate repair being the standard of care. This philosophy arose from Parmley’s 1958 study documenting a death rate at the scene of up to 85% and a subsequent mortality rate in nonoperated survivors of 1% per hour for the first 48 hours. However, this report was an autopsy study, reflecting the natural history of the worst injured as well as the natural history of no treatment, specifically blood pressure control. Currently, while the scene mortality remains similar, it appears that in patients who survive to admission, 1/3 (roughly 5% of the whole) present unstable or become unstable acutely within a short period. The mortality in this group approaches 100%. The remaining 2/3 (10% of the whole) remain stable, and if treated appropriately, the mortality is approximately 25%, with the primary cause of death being associated injuries. The mortality of patients who present with a systolic blood pressure of <90 mm Hg or who drop their pressure to below this level within 1 hour of admission is approximately 70%, while for those who remain stable, mortality is approximately 20%.

In the vast majority of cases, the diagnosis is suggested by plain chest radiographic (CXR) evidence of mediastinal blood. Although sternal and first rib fractures have been used as criteria for angiography, they have very low association with thoracic aortic rupture and in isolation do not warrant routine angiography. When combined with a high degree of clinical suspicion based on mechanism, plain CXR has a ≥98% sensitivity, although specificity can be as low as 10% to 45%. Up to 7% of patients with aortic rupture have normal CXR initially. These patients may present over the ensuing hours or days with gradually increasing mediastinal shillouttes. Thoracic aortography has a sensitivity of nearly 100% and specificity of 98%. False negative studies have been attributed to small intimal lesions and false positives to atheromatous plaques and/or anatomic variants. Computed tomography (CT) scanning has been advocated as a “screening tool” but recently has been superseded by CTA (angiography), which has a sensitivity and specificity approaching angiography. In addition, 3-D reconstructions give valuable data on which to plan operative or endovascular approaches. If both modalities are available, angiography is ideally used when it is required for another reason (such as pelvic embolization, concern for cerebrovascular injury, and so on), while CTA can be used if there is indication for CT scan (to assess the abdomen, for example) or as the primary workup after CXR. Transesophageal echocardiography (TEE) has sensitivity and specificity of 57% to 63% and 84% to 91%, respectively. Due to tracheal obstruction, TEE has limited resolution in the area from the proximal arch to the region between the left common carotid and left subclavian. Advantages of TEE include the ability to be performed during laparotomy, obviating the need for further workup, concomitant assessment of cardiac function, and the ability to discriminate between ulcerated plaques
and true injuries. Intravascular ultrasound (IVUS) has been used in a similar fashion. Both modalities appear to have their greatest utility in assessing equivocal findings on CTA or angiography, for following lesions that are not deemed immediately operable, and in the case of IVUS, assisting in placing stent grafts.

When the diagnosis is suspected on the basis of CXR, immediate control of blood pressure is critical. The target pressure has been described as “less than 120 mm Hg” but more recently it has been argued that a pressure “lower than admission” is sufficient to significantly reduce the risk of rupture during workup. Short-acting beta-blockers, such as esmolol or labetalol, are excellent agents. Pure vasodilating agents (Nipride principally) are not favored, as the reflex increased heart rate increases ΔP/ΔT and may aggravate spinal ischemia by causing shunting of blood away from the cord. Pain control is often all that is required to return blood pressure to acceptable levels. It is likewise an important endpoint in assessing the adequacy of treatment.

Nearly 3/4 of patients have significant associated injuries. The management and prioritization of treatment of associated injuries can be difficult, but in general patients who are hypotensive are more likely to be unstable because of associated injuries (principally pelvic and/or intra-abdominal hemorrhage). Patients who have a grossly positive diagnostic peritoneal lavage (DPL) should undergo laparotomy first, while stable patients with only count positive DPL should have the aortic injury addressed first. Likewise, in patients who have CT evidence of an intra-abdominal injury that could account for the instability, a laparotomy should be performed first. These are difficult decisions.

The vast majority of patients who present with stable blood pressure and are immediately started on beta-blockade remain stable. However, patients with either a hemothorax >500 cc without pneumothorax, supraclavicular hematoma, and/or “pseudocoarctation” are at high risk of early free rupture and should be operated upon immediately unless there are significant contraindications.

Because of associated injuries, 20% to 50% of patients may not be candidates for immediate operative repair (Table 79-2). As noted previously, the cornerstone of therapy becomes careful blood pressure control, or “hypotensive” therapy. Whether or not the risk of free rupture during this period can be determined by the extent of injury is not clear. The risk of rupture when beta-blockade is possible appears to be less than 5%, but it can occur even with minimal injury. The risk of expansion or rupture is greatest in the first 5 to 7 days, after which the natural history of aortic rupture is similar to that of nontraumatic aneurysmal disease, presumably due to the secondary fibrotic reaction. Serial studies (such as helical CTA) every 48 to 72 hours for the first 7 days can be used to follow the lesion and assure stability. Evidence of growth would prompt earlier intervention, even if the risk is greater.

“Hypotensive” therapy can itself be associated with complications. Patients with closed head injury and elevated intracranial pressure may have cerebral perfusion pressure affected, leading to secondary brain injury. These patients may be better treated aggressively to allow the cerebral perfusion pressure to be “driven.” Prolonged lower pressure may result in end organ dysfunction. Thus, urgent repair should still be considered the standard of care, unless there are specific indications to delay surgery.

In the vast majority of cases, a posterolateral 4th intercostal space approach provides the best exposure and access. Lower incisions will not allow access to the root of the left subclavian artery. The left lung must be able to be collapsed, and this is usually achieved with a double lumen endotracheal tube, although newer endobronchial blockers that allow suctioning can be tried if airway edema or other issues contraindicate changing from a single lumen to a double lumen tube.

Proximal exposure requires an appreciation of the likelihood that a tear extends proximal to or beyond the origin of the left subclavian artery (LSCA) into the arch. Lesions within 1 cm of the LSCA origin pose specific anatomic concerns. A number (estimated 14% in one review) will have occult proximal extension or a separate tear, such that clamping distal to the origin of the LSCA would not allow operative correction and might lead to acute aortic disruption. This emphasizes the importance of obtaining aortic control proximal to the LSCA in any case where there is a doubt as to the proximity of the tear to the vessel. Tears close to LSCA are also associated with an increased risk of rupture during proximal dissection, possibly due to a combination of factors: proximal extension already noted; larger size of tears; and inadvertent dissection distally along the medial aspect of the arch entering the injury site. The airway lies immediately behind the aorta at this point and can complicate dissection. In addition, the exposure of these more proximal injuries is slightly more difficult, leading to longer cross-clamp times. Compounding these issues, proximal dissection requires mobilization of the vagus nerve proximal to the point that the recurrent nerve originates, leading to a greater incidence (10% to 20%) of vocal cord paralysis. It is advisable, in patients who are not actively bleeding, to institute bypass first, then perform distal exposure and mobilize the subclavian artery, leaving the proximal exposure to the last so that if bleeding occurs, everything is ready for repair.

A variety of techniques have been described, ranging from graft interposition, to resection and end-to-end anastomosis, to patch repair, to primary repair. In as many as 50% of cases (depending on the series), primary repair (construed as either end-to-end reconstruction or debridement followed by re-approximation of the injured portion of the vessel) has been performed, and the argument in favor of this is shorter cross-clamp times and reduced risk of prosthetic graft infection. If a graft
is used, sizing the graft based on the distal aortic diameter but then trimming the proximal portion of the graft at an angle while laying it out so that the graft lies in a proper attitude will prevent angulation, distortion, and problems with oversizing the graft.