Imaging assessment of penetrating injury of the neck and face

Historically, by the end of World War II, the unacceptably high mortality rate related to penetrating neck wounds (because of associated neurovascular injury going unrecognised) prompted a shift in the management of such wounds towards mandatory exploration of all surgically accessible injuries that penetrated the platysma [3‐7]. It was much later that Monson et al. and Roon and Christiensen [3, 7, 8] attempted to standardise reporting of penetrating neck injuries and improve outcomes by dividing the neck into three anatomic zones of injury. These three anatomical zones are described as follows: zone 1 extends from the level of the clavicles and sternal notch to the cricoid cartilage; zone 2, the largest and most exposed area, extends from the level of the cricoid cartilage to the angle of the mandible; zone 3 extends from the level of the angle of the mandible to the base of the skull (Fig. 1). Traditionally, zone 1 and zone 3 injuries were largely managed conservatively because of difficulties relating to surgical access, with surgery reserved for very few, selected cases . Studies of civilian experience of penetrating neck injury in the latter part of the twentieth century by groups such as Monson et al. revealed that the mandatory exploration of the surgically accessible zone 2 yielded high rates of negative explorations [3, 9‐21]. This has led to a current modern-day management paradigm of selective neck exploration based on clinical and radiological criteria. The emergence of multi-detector computed tomography (MDCT) technology has facilitated the use of CT angiography (CTA) as a reliable, sensitive, fast and non-invasive radiological imaging modality in the assessment of penetrating neck injuries [18, 19, 21‐24] (Fig. 2a, b). CTA has allowed detection of direct and indirect signs of vascular injury not only in zone 2 but also in zone 1 and zone 3; it has also allowed valuable radiological assessment of non-vascular structures in the neck such as the upper aerodigestive tract and cervical spine. CTA has been shown to significantly reduce the number of negative surgical neck explorations [3, 5, 18, 19, 22‐26].

With the significant overall reduction in surgical neck exploration in patients sustaining penetrating neck injury, the use of the zonal descriptive classification has fallen out of favour to a degree; improved cross-sectional CT imaging of such injuries also reveals that, despite the identification of an entry wound in one zone, the track of the penetrating injury may traverse other zones. Nevertheless, the authors still favour the use of such descriptive classification to assist the treating trauma surgeon and recommend familiarity with the classification by any reporting radiologist. It is worth noting that a proportion of patients with penetrating neck injury will require immediate surgical intervention without radiological assessment: current indications for such immediate surgical intervention include: those patients with an expanding haematoma, haemorrhagic exsanguination, shock, airway compromise and massive subcutaneous emphysema [27] (Fig. 3a–c). Nevertheless, the number of patients requiring MDCT angiographic assessment for penetrating neck injury may be considerable in any busy trauma unit. The authors audited the number of patients undergoing emergency CT imaging for projectile and non-projectile penetrating trauma of any part of the body in a 12-month period at their institution (a total of 456 cases): the number of CT angiograms undertaken for non-projectile penetrating neck injury in a 12-month period was also evaluated as part of this audit and 93 such patients were identified. Of these, breach of platysma was demonstrated on multi-detector CT angiography (MDCTA) in 72 %.

Description of the detailed anatomy of the neck is beyond the scope of this review. However, the most salient considerations include the distinctive fascial envelopes in the neck. The neck is invested in two fascial layers: the superficial fascia and the deep cervical fascia. The superficial fascia overlies the platysma and forms part of the superficial musculo-aponeurotic system in the face; the deep cervical fascia is made up of the investing layer, pretracheal layer, prevertebral layer and carotid sheath. Penetrating injuries confined to the superficial fascia—that is, injuries not breaching the platysma—are not life-threatening.

The mechanisms by which projectiles, such as those discharged from firearms (bullets, for example), cause tissue injury (terminal ballistics) are crushing and stretching—that is, shockwave formation and temporary cavitation [28‐30]. Harvey and Korr [30] described two types of tissue damaging pressure waves produced by projectiles such as bullets: the first wave, the sonic shock wave, relates to the sound of the projectile striking the target; it transmits at the speed of sound (faster than the projectile itself) and no temporary cavity is associated with this sonic shock wave; the secondary pressure wave, the temporary cavity, is formed when the penetrating projectile strikes and deforms tissue and radiates out laterally, propagating tissue damage.

The lethality of a projectile relates significantly and directly to its deformability and the degree of fragmentation it undergoes in the target. The Hague Convention of 1899 has strict specifications for military-use bullets in order to limit the lethality of such ammunition; true full-metal jacketed military bullets resist deformation and fragmentation and are in fact designed to wound but not to kill (Fig. 4a). The harsh reality of warfare philosophy is that the soldier wounded but not killed by a bullet on the frontline is a far greater drain on the economical and manpower resources of the enemy than is a fatality. Clearly, the construction of some projectiles such as Improvised Explosive Devices (IED) previously and currently employed in certain international conflicts by certain factions do not adhere to the Hague Convention and the injury potential of such devices is well recognised. Similarly, civilian (and hunting) ammunition does not adhere to the restrictions of the Convention and is therefore capable of causing tremendous tissue damage [28, 29].

In general, bullet wounds (and, in fact, any projectile wound) are more severe when the projectile yaws through tissue, when the projectile fragments or deforms (into a mushroom shape, for example), when the projectile is large, or when the projectile is travelling at high velocity [28]. One of the “recommendations” of the hollow-point and soft-point bullet (the former are used by specialist fire-arms officers in the UK) is that they are less likely to exit the target, which poses smaller risk to bystanders. This is because they exhibit significant deformation as a result of their construct, transferring most or all of their kinetic energy to the tissue and usually stopping within the body target. The lethality of the hollow-point and soft-point bullet (Fig. 4b) is therefore very high and it is for this reason that this type of ammunition is also used in hunting, supposedly expediting death of the target animal as rapidly and as humanely as possible.

The terminal ballistics of shotgun injuries warrant special consideration: shotgun pellets create tissue damage proportional to the distance from the target, proportional to the number of pellets that hit the target, and proportional to the size and pattern of the pellet strike. At close range, shotguns inflict massive tissue damage and large avulsion wounds. However, this type of weapon is less effective at greater distances from the target, with projectile velocity falling considerably beyond 20–40 m.

MDCT scanning is now widely available and MDCTA is the recommended primary imaging modality and method of choice. This allows not only vascular assessment but also appropriate radiological evaluation of the osseous cervical spine and skull base, of the craniofacial skeleton and valuable assessment of the viscera and fascial planes of the neck and face. At the authors’ institution, both 64- and 128-MDCT imaging (Siemens) is available but all such scans are now currently performed at 128-slice capability. CT protocol parameters currently in use at the authors’ institution are described in brief: a collimation of 128 × 0.6 mm (for the 128-detector scanner) or 64 × 0.6 mm (for the 64-slice scanner); 120 kV is operated on both types of scanner and a reference mAs of 220 on the 64-slice scanner and 120 on the 128-slice scanners; Siemens care kV is utilised on the 128-slice scanners and CareDose4D is utilised on all scanners; 100 ml intravenous contrast is injected for CT angiographic assessment of the neck vessels; where additional CT angiographic assessment of the intracranial vessels is required, a 50 ml normal saline “chaser” is injected following the intravenous contrast injection. A contrast injection rate of 5 ml per second through (at smallest) a 19-G venous cannula is used. Bolus-tracking from the arch of the aorta is routinely operated with a delay of 7 s after triggering of contrast injection (initiated at 100 Houndsfield units) on the 128-slice scanners but a delay of 6 s on the 64-slice scanner.

Liberal use of multiplanar reformats, always in correlation with the axial source dataset images, is advised. Appropriate window-level and window-width settings for the different anatomical regions (for example, the osseous cervical spine and skull base) should also be utilised.

Vascular injuries are the most common injuries associated with penetrating neck trauma, estimated to occur in up to 40 % of patients [27] and with arterial injury representing between 15 % and 25 % [19, 24, 25, 31]. Of arterial injuries, 80 % involve the carotid vessels and up to 43 % involve the vertebral arteries [19, 33, 34]. Morbidity and mortality relate to complicating neurological deficits as a result of stroke (Fig. 3a) and not simply as a result of exsanguination. Physical findings caused by neurological deficits can include Horner’s syndrome or cranial nerve dysfunction (Fig. 5a, b) as well as overt anterior or posterior circulation stroke.

MDCTA accurately demonstrates the vascular injuries associated with projectile and non-projectile penetrating trauma of the neck. These include: pseudoaneurysm formation, arteriovenous fistula formation (Fig. 6), vessel transection, intimal injury with flap formation, dissection, active bleeding, and partial or complete occlusion of the carotid or vertebral arteries (Figs. 3b and 7a, b). The involvement of the internal jugular vein by penetrating injury is not infrequently identified on CTA and for this reason, in the authors’ experience, some early venous contamination on the MDCT acquisition is not always deleterious. Injury to venous structures should not be ignored: venous injuries occur in nearly 20 % of patients with penetrating trauma of the neck and are frequently missed at physical examination [35].

Studies comparing the accuracy of CTA (single-slice and multi-detector) with catheter angiography have prompted a shift of emergency vascular imaging assessment of appropriate penetrating neck injuries to MDCTA as the primary imaging modality and technique of choice [3, 18, 20‐22, 27, 36]. The value of MDCTA in penetrating neck injury assessment of non-vascular injuries affecting the upper aerodigestive tract and osseous structures has also been consolidated [26].

The facility of catheter angiography in the management of penetrating vascular injuries of the neck should not be dismissed in its entirety. Although no longer the first-line diagnostic imaging modality of choice, catheter angiography may still be required as part of the management of certain arterial injuries identified on MDCTA in the haemodynamically stable patient amenable to therapeutic radiological endovascular intervention. In addition, particularly in the setting of projectile penetrating injury, where streak artefact from metallic fragments such as bullet fragments or shrapnel may, on occasion, hinder CT interpretation, catheter angiography may be indicated. An example of such a scenario might be where wound trajectory in relation to major vessels is of significant-enough concern, but CT interpretation is significantly hindered by streak artefact. Magnetic resonance (MR) imaging, and more specifically MR angiography, has no role in the imaging assessment of metallic projectile penetrating neck injury because of the presence of potentially ferromagnetic foreign body material. MR has only a very limited and selective role in the acute imaging assessment of non-projectile penetrating neck injury: where non-projectile penetrating injury to the cervical spine or craniocervical junction is suspected clinically, MR imaging will be essential [28], particularly as these patients frequently survive such injuries despite severe morbidity.

The use of ultrasound imaging in penetrating neck injury is very limited when compared with MDCTA, particularly given its inability to evaluate the upper aerodigestive tract and spine.

Penetrating trauma to the face may carry significant morbidity and mortality, not least because of the associated intracranial component of injury which may accompany such trauma. (This component will not be discussed further in this review). It can result from projectile penetrating injury, most typically gunshot injuries (and less typically blast injuries), or from non-projectile mechanisms. Direct gunshot wounds to the face encountered in the civilian setting are not uncommonly self-inflicted, in which case the weapon of choice tends to be the shotgun. The imaging modality of choice remains non-contrast CT of the facial bones and brain. In certain situations, MDCTA may also be appropriate. The reporting radiologist should be aware of the substantial secondary tissue damage caused by in-driven fragments of bone and teeth, which themselves become secondary projectiles (Figs. 12a–c and 13a, b). Shotgun pellets at short range cause significant propagation of tissue damage as a result of the “billiard-effect” of the individual pellets [28, 29, 42, 48‐51]. In addition, the potential for shotgun pellets, bullet fragments or native organic tissue such as fragmented bone and teeth to enter the oral cavity, migrate and be aspirated, should also be considered by the radiologist when assessing the emergency imaging (Fig. 14a–c). There has been no documented evidence of lead poisoning caused by swallowed lead-containing bullets or shotgun pellets.

Non-projectile penetrating facial trauma most often results from injury by sharp implements such as knives, glass or screwdrivers, as well as accidental sporting or occupational impalement injuries caused by metal railings, garden materials such as sharp wooden poles and fencing or tree branches. The ability of such objects to penetrate deeply to the parapharyngeal and retropharyngeal compartments at the skull base where vital structures such as the internal carotid arteries course and may be at risk, should be considered. The possibility of more local superficial soft tissue damage such as parotid parenchymal or parotid duct injury should also be considered. Damaged external carotid artery branches can be a source of catastrophic haemorrhage, which may warrant emergency endovascular treatment. MDCTA may be of significant value in such circumstances. Organic material such as impaled wood fragments can be somewhat elusive on CT assessment unless appropriate window settings are employed, depending on the amount of air and fluid within the interstices of the wood [52‐54]. Impaled wood fragments may appear as “air” on CT and MRI assessment, but the radiologist should be alert to the possibility of an embedded fragment of wood if this “air” exhibits a geometric margin (Fig. 15).