Chapter 2 Airway Imaging
Principles and Practical Guide
Interpretation of radiologic studies is not usually in the domain of anesthesiologists. However, imaging studies can provide a wealth of information regarding the airway. This information can be aptly used for formulating an anesthetic plan. Currently, radiology is not part of the curriculum of any anesthesia residency training program; for this reason, most anesthesiologists have only rudimentary skills in the interpretation of radiologic studies. The main goal of this chapter is to introduce anesthesiologists to normal airway anatomy as visualized on conventional radiography (plain x-ray films) and on cross-sectional imaging such as computed tomography (CT) and magnetic resonance imaging (MRI) and to illustrate the anatomic variants and pathologic processes that can compromise the airway. The technical aspects of each imaging modality are reviewed briefly. Emphasis is placed on evaluation of the airway using available radiologic studies, which are most often performed for non-airway issues. Relevant information regarding the airway is readily available and ranges from conventional chest radiographs to high-resolution cross-sectional imaging of the neck or chest by CT or MRI. The clinical examples in this chapter focus on the pathologic processes involving the airway that are most relevant to anesthesiologists and include short discussions of some of the more common abnormalities.
Simplistically, the airway can be regarded as a tubular conduit for air inhaled from the nares to the tracheobronchial tree. The soft tissue structures bordering the airway have warranted more of the radiologist’s attention. The integrity of the airway with its natural contrast is usually referenced with respect to extrinsic impression, compression, encroachment, or displacement. Segmentation of the airway into compartments (i.e., head and neck, chest) is artificial and is usually done only for the ease of discussion. However, imaging of this airspace as a unique entity is gaining popularity. Knowledge of the technical differences among imaging modalities can aid in ordering and interpreting the imaging study. This is especially important when selecting a study that will best depict the anatomic structures and pathologic processes of the airway.
Wilhelm Conrad Roentgen, a German physicist, discovered x-rays on November 8, 1895, while studying the behavior of cathode rays (electrons) in high-energy cathode ray tubes. By serendipity, he noticed that a mysterious ray that escaped the cathode tube penetrated objects differently, and he named this the x-ray. For his work, he was awarded the first Nobel Prize for Physics in 1901.1
X-rays are a type of electromagnetic radiation; as the name implies, they transport energy through space as a combination of electric and magnetic fields. Other types of electromagnetic radiation include radio waves, radiant heat, and visible light. In diagnostic radiology, the predominant energy source used for imaging is ionizing radiation (i.e., alpha, beta, gamma, and x-rays). The science of electromagnetic waves and x-ray generation is very complex and exceeds the scope of this text. In principle, x-rays are produced by energy conversion as a fast stream of electrons is suddenly decelerated in an x-ray tube.2 The localized x-ray beam that is produced passes through the part of the body being studied. The final image is dependent on the degree of attenuation of the beam by matter.
Attenuation is the reduction in the intensity of the beam as it traverses matter of different constituents. It is caused by absorption or deflection of photons from the beam. The transmitted beam determines the final image, which is represented in shades of gray.3–5 The lightest or brightest area on the film or image represents the greatest attenuation of the beam by tissue and the least amount of beam transmitted to film. For example, bone is a high-density material that attenuates much of the x-ray beam; images of bone on x-ray films are very bright or white. A plain film image is a one-dimensional collapsed or compressed view of the body part being imaged. This information can also be presented in a digital format without the use of traditional x-ray films.
Compared with other, more sophisticated imaging modalities, conventional radiography has limited range in the display of tissues of different density and spatial resolution. Its advantages are lower cost of the examination, overall lower radiation exposure compared with a more comprehensive CT examination, and presentation of anatomy with a larger field of view. The head, chest, abdomen, or extremity can be visualized on a single film or digitized image, and therefore the image appears more familiar to nonradiologists. Also, the plain x-ray film can be obtained quickly at the patient’s bedside in any location in the hospital. The combination of x-radiography with cine mode allows radiologists to obtain dynamic images, which are used to evaluate organ function (e.g., barium swallow to evaluate deglutition, intravenous pyelogram to assess renal function, vascular studies).
After the discovery of x-rays, it became apparent that images of the internal structures of the human body could yield important diagnostic information. However, the usefulness of x-ray studies is limited because they project a three-dimensional (3-D) object onto a two-dimensional display. With x-rays, the details of internal objects are masked by the shadows of overlying and underlying structures. The goal of diagnostic imaging is to bring forth the organ or area of interest in detail while eliminating unwanted information. Various film-based traditional tomographic techniques were developed, culminating in the creation of computerized axial tomography or computed tomography (CT).6 The first clinically viable CT scanner was developed by Hounsfield and commercially marketed by EMI Limited (Middlesex, England) for brain imaging in the early 1970s.7 Since then, several generations of CT scanners have been developed.
As with conventional plain film radiography, CT technology requires x-rays as the energy source. Whereas conventional radiography employs a single beam of x-rays from a single direction and yields a static image, CT images are obtained with the use of multiple collimated x-ray beams from multiple angles, and the transmitted radiation is counted by a row or rows of detectors. The patient is enclosed in a gantry, and a fan-shaped x-ray source rotates around the patient. The radiation counted by the detectors is analyzed with the use of mathematical equations to localize and characterize tissues by density and attenuation measurements. A single cross-sectional image is produced with one rotation of the gantry.6 The gantry must then “unwind” to prepare for the next slice while the table carrying the patient moves forward or backward by a distance that is predetermined by slice thickness. An intrinsic limitation of this technique is the time required for moving the mechanical parts.
The introduction of slip-ring technology in the 1990s and the development of faster computers, high-energy x-ray tubes, and multidetectors enabled continuous activation of the x-ray source without having to unwind the gantry and also allowed continuous movement of the tabletop. This process, known as helical CT, is used in the latest generation of CT scanners. Because the information acquired using helical CT is volumetric, in contrast to the single slice obtained with conventional CT, the entire thorax or abdomen can be scanned in a single breath-hold. Volumetric information makes it possible to identify small lesions more accurately and allows better 3-D reformation. Because of the higher speed of data acquisition, misregistration and image degradation caused by patient motion are no longer significant concerns. This is especially important when scanning uncooperative patients and trauma victims. The absorbed radiation dose used in multidetector helical CT (as compared with conventional single-detector row CT) is dependent on the scanning protocol and varies with the desired high-speed or high-quality study.8
Practically speaking, CT examinations have become routine. The spatial resolution of CT is the best of all the imaging modalities currently available. The advantage of CT technology is that it can depict accurately any pathology involving bones. Data acquisition is very quick. CT can be used to produce images in all three planes and to provide information for surface rendering and 3-D reformation, which allows the display of organs in an anatomic format that can be easily recognized by clinicians.
MRI has become one of the most widely used imaging modalities in diagnostic radiology. In contrast to conventional radiography and CT, MRI uses no ionizing radiation. Instead, imaging is based on the resonance of the atomic nuclei of certain elements such as sodium, phosphorus, and hydrogen in response to radio waves of the same frequency produced in a static magnetic field environment. Current clinical MRI units use protons from the nuclei of hydrogen atoms to generate images because hydrogen is the most abundant element in the body. Every water molecule contains two hydrogen atoms, and larger molecules, such as lipids and proteins, contain many hydrogen atoms. Powerful electromagnets are used to create a magnetic field, which influences the alignment of protons in hydrogen atoms in the body. When radio waves are applied, protons are knocked out of natural alignment, and when the radio wave is stopped, the protons return to their original state of equilibrium, realigning to the steady magnetic field and emitting energy, which is translated into weak radio signals. The time it takes for the protons to realign, referred to as a relaxation time, is dependent on the tissue composition and cellular environment.9 The different relaxation times and signal strengths of the protons are processed by a computer, generating diagnostic images. With MRI, the chemical and physical properties of matter are examined at the molecular level. The relaxation times for each tissue type, designated T1 and T2, are expressed as constants at a given magnetic field strength. Imaging that optimizes T1 or T2 characteristics is referred to as T1-weighted or T2-weighted imaging, respectively. Tissue response to pathologic processes usually includes an increase in bound water (edema), which lengthens the T2 relaxation time and appears as a bright focus on T2-weighted images.9
MRI is more sensitive, but not necessarily more specific, in detecting pathology than CT, which depicts anatomy with unparalleled clarity. Imaging with MRI provides metabolic information at the cellular level, allowing one to link organ function and physiology to anatomic information. MRI and CT technologies also have other differences: (1) MRI shows poor bony detail, whereas CT provides excellent images of bony structures; (2) hemorrhage, especially if acute, is clearly visible on CT scans but may be difficult to diagnose with MRI because the appearance of blood varies temporally to the evolution of the breakdown products of hemoglobin; and (3) MRI is very susceptible to all types of motion artifacts, ranging from a patient’s movement, breathing, swallowing, and phonation to vascular and cerebral spinal fluid pulsation and flow.
MRI scanners operate in a strong magnetic field environment, and strict precautions must be observed. Any item containing ferromagnetic substances that is introduced into the magnetic field environment can become a projectile and result in deleterious consequences for patients, personnel, and the MRI scanner itself. Therefore, no metal objects should be brought into the MRI suite if one is not absolutely certain about their composition. Only specially designed nonferromagnetic equipment is used in the MRI suite, including anesthesia machines, monitoring equipment, oxygen tanks, poles for intravenous equipment, infusion pumps, and stretchers. Pagers, telephones, handheld organizers and computers, credit cards, and analog watches must also be removed, because the strong magnetic field can cause malfunction or permanent damage to them. Patients must be carefully screened for implantable pacemakers, intracranial aneurysm clips, cochlear implants, and other metallic foreign objects before entering the MRI environment.
To illustrate the usefulness of conventional radiography in evaluating the airway, this discussion focuses on the interpretation of plain films of the cervical spine, chest, and neck. Before the CT era, these were probably the most frequently ordered x-ray studies in the hospital setting, and they are ubiquitous in patients’ film jackets and on picture archiving communication systems (PACS). As a composite, these studies provide a picture of the entire airway. Although these radiologic studies are usually obtained for reasons other than airway evaluation, it is actually in the group of patients who are “normal” or “cleared for surgery” that one may glean important observations about the airway. The anatomy and pathology displayed by plain film radiographs may alert the anesthesiologist to potential difficulties in securing the patient’s airway and help him or her to develop an alternative anesthetic plan. In this sense, the information about the airway that is inherent to these x-ray examinations is gratuitous. The following sections address the basics of plain film interpretation with respect to imaging of the airway anatomy and pathology.
The cervical spine connects the skull to the trunk; it articulates with the occiput above and the thoracic vertebrae below. The bony elements, muscles, ligaments, and intervertebral discs support and provide protection to the spinal cord. On plain films, one can appreciate the bony morphology of the vertebrae and the disc spaces and assess the alignment of the vertebral column very quickly. This indirectly provides information regarding the integrity of the ligaments, which are crucial in maintaining alignment of the cervical spine. However, individual ligaments and muscle groups all have the same or similar attenuation and cannot be differentiated from one another on plain film. A systematic approach is recommended to evaluate the spine for bony integrity, alignment, cartilage, joint space, and soft tissue abnormalities. The disadvantages of cervical spine radiography are the limited range of tissue attenuation and the loss of spatial resolution caused by overlapping bone structures.
The most common indications for obtaining cervical spine radiographs in today’s medical practice are for the evaluation of trauma, spinal stability, and cervical spondylosis and in the search for radiopaque foreign bodies. Different views of the cervical spine are tailored to each clinical need. The most common views are the lateral, anteroposterior (AP), open-mouth odontoid, oblique, and pillar views (Fig. 2-1). In acute cervical spine injury, cross-table lateral, AP, and open-mouth odontoid views are recommended. A lateral view reveals the majority of injuries (Fig. 2-2); however, patients who are rendered quadriplegic by severe ligamentous injuries may demonstrate a normal lateral cervical spine radiograph. When the AP and then the open-mouth odontoid views are added to the cross-table lateral view of the cervical spine, the sensitivity of detecting significant injury is increased from 74% to 82% and then to 93%.10 In today’s practice, cross-sectional imaging (i.e., CT of the spine) has become a mainstay in the evaluation of the cervical spine, especially in the setting of acute trauma. MRI is particularly useful in evaluating the spinal cord.
Figure 2-1 Normal cervical spine series. A, Lateral view shows upper (a) and lower (b) end plates of the third cervical vertebra (C3), transverse process (c), pedicle (d), facet joint (e), articulating facets (f), posterior spinous process (g), posterior arch of C1 (h), anterior arch of C1 (i), atlantoaxial distance (j), and hyoid bone (k). B, Anteroposterior view shows smoothly undulating cortical margins of the lateral masses (a), joint of Luschka (b), superior (c) and inferior (d) end plates, and a midline posterior spinous process (e). C, Open-mouth odontoid view shows the odontoid tip (a), centered between the lateral masses of the axis; symmetrical lateral margins of the lateral atlantoaxial joints (b); and a spinous process (c). D, Oblique view shows laminae of the articular masses (a), reflecting the shingling effect, and an intervertebral (neural) foramen (b).
In brief, a normal lateral cervical radiograph should demonstrate seven intact vertebrae and normal alignment of the anterior and posterior aspects of the vertebral bodies. This is especially important for trauma victims, because 7% to 14% of fractures are known to occur at the C7 or C7-T1 level.11 The posterior vertebral body line is more reliable and must be intact. The anterior vertebral line is often encumbered by the presence of anterior osteophytes. Normal facet joints overlap in an orderly fashion, similar to shingles on a rooftop. The spinolaminar line, which is the dense cortical line representing the junction of the posterior laminae and the posterior spinous process, is uninterrupted. Relative uniformity of the interlaminar (interspinous) distances should be observed. The posterior spinal line (i.e., posterior cervical line), an imaginary line extending from the spinolaminar line of the atlas to C3 (Fig. 2-3), should demonstrate a continuous curve in parallel to the posterior vertebral body line; the distance between the two correlates with the spinal canal diameter.12
Figure 2-3 Normal lateral cervical spine radiograph (A) demonstrates normal alignment. a, Anterior spinal line; b, posterior vertebral line; c, posterior spinal line. B, Lateral scout view of a computed tomography scan demonstrates anterior subluxation of C4 on C5.
The anatomy and integrity of the craniocervical junction are crucial to the anesthesiologist. To achieve successful and safe endotracheal intubation, the anterior atlantodental interval (AADI), the vertical and anterior-posterior position of the dens, and the degree of extension of the head on the neck must be considered. The anterior arch of C1 bears a constant relationship to the dens; this is the AADI or predental space. It is defined as the space between the posterior surface of the anterior arch of C1 and the anterior surface of the dens. In flexion, because of the physiologic laxity of the cervicocranial ligaments, the anterior tubercle of the atlas assumes a more normal-appearing relationship to the dens, and the AADI increases in width, greater rostrally than caudally. In children and with flexion in adults, the AADI is normally about 5 mm. In adults, it is generally accepted that the AADI is 3 mm or less (Fig. 2-4).12
Figure 2-4 Anterior atlantodental interval (AADI). A, Lateral radiograph of the cervical spine in an adult patient. The AADI or predental space (arrows) is normally less than 3 mm. B, Lateral radiograph of the cervical spine in a pediatric patient. An AADI of up to 5 mm (curved arrow) can be normal in a child. The basion, the midpoint of the anterior border of the foramen magnum, is indicated by the straight arrow. The dotted line is an imaginary line indicating the inferior extension from the basion, and the posterior axial line is indicated by the solid line. The distance between the dotted line and the solid line is the basion-axial interval (BAI); it should be 12 mm or less for a normal occipitovertebral relationship in a child.
The bony structures of the atlantoaxial joint provide mobility (e.g., rotational movement) rather than stability. Therefore, the ligaments play a significant role in stability. The most important ligaments in the upper cervical spine are the transverse ligament, the alar ligaments, and the tectorial membrane. If the transverse ligament is disrupted and the alar and apical ligaments remain intact, up to 5 mm of movement at the atlantoaxial joint can be seen.13 If all the ligaments have been disrupted, the AADI can measure 10 mm or larger. In atlantoaxial subluxation, the dens is invariably displaced posteriorly, which causes narrowing of the spinal canal and potential impingement of the spinal cord. The space available for the spinal cord is defined as the diameter of the spinal canal as measured in the AP plane, at the C1 level, that is not occupied by the odontoid process. In the normal spine, this space is approximately 20 mm.13
Pseudosubluxation and pseudodislocation are terms applied to the physiologic anterior displacement of C2 on C3 that is frequently seen in infants and young children (Fig. 2-5). Physiologic anterior displacement of C2 on C3 and of C3 on C4 occurs in 24% and 14%, respectively, of children up to 8 years of age.14
Figure 2-5 Pseudosubluxation at C2-C3. T2-weighted sagittal magnetic resonance cervical spine study demonstrates physiologic anterior displacement of C2 on C3 in a child. Also seen are normal soft tissue masses encroaching on the airway from adenoids (a), palatine tonsils (b), and lingual tonsils at the base of the tongue (c).
In pediatric trauma cases, if C2 is anteriorly displaced and there are no other signs of trauma such as posterior arch fracture or prevertebral soft tissue hematoma, the spinolaminar lines of C1 through C3 should have a normal anatomic relationship. In a neutral position, the spinolaminar line of C2 lies on or up to 1 mm anterior or posterior to the imaginary posterior spinal line. If the C2 vertebra is intact, as the C2 body glides forward with respect to C3 during flexion, the spinolaminar line of C2 moves 1 to 2 mm anterior to the posterior spinal line. Similarly, with extension, the posterior translation of the C2 body is mirrored by similar posterior displacement of the spinolaminar line of C2 with respect to the posterior spinal line.
In traumatic spondylolisthesis, which is rare in children but more common in adults, the C2 body would translate anteriorly in flexion and posteriorly in extension, and the posterior spinal line would be maintained because of intact ligaments. However, flexion and extension films are not advisable if traumatic spondylolisthesis is suspected.
Important to rigid laryngoscopy and endotracheal intubation is the distance between the occiput and the posterior tubercle of C1, known as the atlanto-occipital distance (Fig. 2-6), which is quite variable from individual to individual. Head extension is limited by the abutment of the occiput to the posterior tubercle of C1. It has been proposed that a shorter atlanto-occipital distance decreases the effectiveness of head extension and contributes to difficult intubation.13,15 Occipitalization of C1 with the occiput (atlanto-occipital fusion) not only limits head extension but also adds stress to the atlantoaxial joint. Although the majority of head extension occurs at the atlanto-occipital joint, some extension can also occur at C1-C2.15 Nichol and Zuck observed that in patients with limited or no extension possible at the atlanto-occipital joint, general extension of the head actually brings the larynx “anterior,” thus limiting the visibility of the larynx on laryngoscopy.15
Figure 2-6 Atlanto-occipital distance. A, Lateral radiograph of the cervical spine in neutral position; the arrow indicates the atlanto-occipital distance. B, Lateral radiograph of the cervical spine in hyperextension. Head extension is limited by the abutment of the occiput to the posterior tubercle of C1.
Ossification of the atlas begins with the lateral masses during intrauterine life. At birth, neither the anterior nor the posterior arches are fused. Fusion of the anterior arch is complete between 7 and 10 years of age. During the second year, the center of the posterior tubercle appears, and by the end of the fourth year, the posterior arch becomes complete.12 Nonfusion of the anterior or the posterior arch, or both, exists as a normal variant in adults and should not be mistaken for fracture (Fig. 2-7).
The second cervical vertebra, the axis (C2), is the largest and heaviest cervical segment. The C2 vertebra arises from five or six separate ossification centers, depending on whether the centrum has one or two centers. The vertebral body is ossified at birth, and the posterior arch is partially ossified. They fuse posteriorly by the second or third year of life and unite with the body of the vertebra by the seventh year.
The odontoid process (dens) serves as the conceptual body of C1, around which the atlas rotates and bends laterally. In contrast to the other cervical vertebrae, C2 does not have a discrete pedicle. The dens is situated between the lateral masses of the atlas and is maintained in its normal sagittal relationship to the anterior arch of C1 by several ligaments, the most important of which is the transverse atlantal ligament. Superiorly, the dentate (apical) ligament extends from the clivus to the tip of the dens. Alar ligaments secure the tip of the dens to the occipital condyles and to the lateral masses of the atlas. They are the second line of defense in maintaining the proper position of the dens. The tectorial membrane is a continuum of the posterior longitudinal ligament from the body of C2 to the upper surface of the occipital bone anterior to the foramen magnum.
The dens ossifies from two vertically oriented centers that fuse by the seventh fetal month. Cranially, a central cleft separates the tips of these ossification centers (Fig. 2-8), and it can mimic a fracture if ossification is incomplete. The ossiculum terminale, the ossification center for the tip of the dens, may be visible on plain films, conventional tomograms, or CT scans and unites with the body by age 11 or 12 years. Failure of the ossiculum terminale to develop or failure to unite with the dens may result in a bulbous cleft dens tip. A nonunited terminal dental ossification center, called the os terminale, may be mistaken for a fracture of the odontoid tip.
The position and anatomy of the dens with respect to the anterior arch of C1 and the foramen magnum are worthy of attention. Congenital anomalies of the odontoid process, such as hypoplasia, can result in a loss of the buttressing action of the dens during extension and subsequent compression of neural elements. Examples of conditions that are associated with odontoid hypoplasia are the Morquio, Klippel-Feil, and Down syndromes; neurofibromatosis; dwarfism; spondyloepiphyseal dysplasia; osteogenesis imperfecta; and congenital scoliosis.13,16 Patients with these conditions are predisposed to atlantoaxial subluxation and craniocervical instability, and hyperextension of the head for intubation should be avoided. In addition, congenital fusion of C2 and C3 (Fig. 2-9), whether occurring as an isolated anomaly or as part of Klippel-Feil syndrome, places added stress at the C1-C2 junction.
Figure 2-9 Congenital fusion of C2 and C3. Lateral cervical spine radiograph (A) and sagittal T1-weighted magnetic resonance (MR) study (B) demonstrate fusion of the C2 and C3 vertebral bodies (dotted arrow) and their lateral and posterior elements (solid arrow). Lateral radiograph (C) and T1-weighted MR cervical spine study (D) of a patient with Klippel-Feil syndrome demonstrate fusion of C2 to C3 and fusion of C4 to C6. Not surprisingly, a disc herniation is present at the point of greatest mobility at C3-C4.
Cervical spine radiographs are obtained for the evaluation of cervical spondylosis (Fig. 2-10). The hypertrophic bone changes associated with this condition are well depicted on radiographic studies. Large anterior osteophytes that project forward may cause dysphagia and difficult intubation. The bone canal and neural foramina are assessed for stenosis; if stenosis is present, precautions can be taken when hyperextending the neck and positioning the patient to avoid exacerbation of baseline neurologic symptomatology. Calcification and ossification are well depicted on radiographic studies.
Ossification of the anterior longitudinal ligament and diffuse idiopathic skeletal hyperostosis have been reported as causes of difficult intubation.17 This can be readily appreciated on plain films. Another condition that may signal difficult intubation is calcification of the stylohyoid ligament (Fig. 2-11).18
Figure 2-11 Calcified stylohyoid ligament. A, Lateral cervical spine radiograph. B and C, Coronal computed tomograms with bone algorithm. Lateral cervical spine flexion (D) and extension (E) views are also shown. Black arrows indicate calcified stylohyoid ligaments, and white arrows indicate the laryngeal ventricle (at the level of the vocal cords). Notice the change in the level of the hyoid bone and vocal cords with flexion and extension of the neck. h, Hyoid bone; s, styloid process; t, calcified thyroid cartilage.
Inflammatory arthropathies involving the atlantoaxial joint with subluxation are classically seen in patients with rheumatoid arthritis or ankylosing spondylitis. However, the underlying causes of atlantoaxial subluxation are quite different in these two entities. Ankylosing spondylitis is characterized by progressive fibrosis and ossification of ligaments and joint capsules. In rheumatoid arthritis, bone erosion, synovial overgrowth, and destruction of the ligaments occur. Patients with rheumatoid arthritis are not only susceptible to AP subluxation at the C1-C2 junction but also at risk for vertical subluxation of the dens. Whether this condition is referred to as “cranial settling,” superior migration of the odontoid process, or basilar invagination, the end result is the same.12 The odontoid process protrudes above the foramen magnum, narrowing the available space for the spinal cord and potentially leading to cord compression with the slightest head extension (Fig. 2-12).13
Figure 2-12 Position of the dens in a normal patient (A), in a rheumatoid patient (B), and in a nonrheumatoid patient with basilar invagination and platybasia (C). A, Postmyelography computed tomogram with sagittal reformation demonstrates normal relationship of the dens with respect to the foramen magnum, brainstem, and anterior arch of C1. A normal atlantoaxial distance (AADI) is seen (arrow). B, T1-weighted sagittal magnetic resonance (MR) study of the cervical spine in a rheumatoid patient with erosion and pannus formation at the atlantoaxial joint resulting in increased AADI (arrow), posterior subluxation of the dens, and brainstem compression. C, Sagittal MR study of the brain in a nonrheumatoid patient shows normal AADI, but basilar invagination and platybasia have resulted in vertical subluxation of the dens and brainstem compression. The line drawn from the hard palate to the posterior lip of the foramen magnum is Chamberlain’s line (dotted line); basilar invagination is defined as extension of the odontoid tip 5 mm or more above this line. Also notice the fusion of the C2 and C3 vertebrae. The small, linear, dark line at the mid-C2 level is the subdental synchondrosis (white arrow).
In response to the effective foreshortening of the spine that occurs secondary to the superior migration of the odontoid process from inflammatory or degenerative disease, there is acquired rotational malalignment between the spine and larynx.19 The larynx and the trachea, because they are semirigid structures and as a result of the tethering effect of the arch of the aorta as it passes posteriorly over the left main bronchus, are predictably displaced caudally, deviated laterally to the left, rotated to the right, and anteriorly angulated. The effective neck length can be affected by superior migration of the dens, severe spondylosis with loss of disc space, or iatrogenic causes secondary to surgery. The soft tissues of the pharynx become more redundant owing to the relative shortening of the neck, which further obscures the view of the larynx. On laryngoscopy, the vocal cords are rotated clockwise. A rotated airway is suspected when the frontal view of the cervical spine demonstrates a deviated tracheal air column.
Historically, bony landmarks other than the spine that can be appreciated on a lateral cervical spine radiograph have been used in the anesthesia arena to preoperatively predict difficult laryngoscopy and endotracheal intubation on the basis of anatomic factors. Mandibular size, the ratios of the various measurements, and their relationship to the hyoid bone have been proposed as predictors of difficult laryngoscopy (Fig. 2-13).20 These measurements are meant to reflect the oral capacity, the degree of mouth opening, and the level of larynx.21,22 It is apparent that the causes of difficult laryngoscopy and endotracheal intubation are multifactorial. Combined with the clinical examination, anatomic measurements and findings assessed by x-ray studies can help alert the anesthesiologist to a potentially difficult airway. In this way, difficult laryngoscopy and endotracheal intubation can be anticipated and not unexpected.
Figure 2-13 Mandibular and hyoid measurements proposed as predictors of difficult laryngoscopy: 1, Anterior depth of mandible; 2, posterior depth of mandible; 3, mandibulohyoid distance; 4, atlanto-occipital distance; 5, thyromental distance. Laryngeal ventricle (solid arrow) demarcates the level of the larynx. The true vocal cords are just below the level of the laryngeal ventricle. e, Epiglottis; h, hyoid bone.
The lateral cervical spine study with bone and soft tissue technique allows an incidental view of the aerodigestive tract and a gross assessment of the overall patency of the airway. Useful ossified cartilage or bony landmarks of the pharynx and larynx that can be appreciated on the lateral neck radiograph are the hard palate, hyoid bone, thyroid, and cricoid cartilages (Fig. 2-14). The hard palate is a bony landmark used to separate the nasopharynx from the oropharynx. The larynx can be thought of as being suspended from the hyoid bone. Muscles acting on the hyoid bone elevate the larynx and provide the primary protection from aspiration. The largest cartilage in the neck is the thyroid cartilage, which along with the cricoid cartilage acts as a protective shield for the inner larynx. The cricoid cartilage is the only complete cartilaginous ring in the respiratory system. It is located at the level where the larynx ends and the trachea begins.
Normal air-filled structures seen on lateral plain films are the nasopharynx, oropharynx, and hypopharynx. Air in the pharynx outlines the soft palate, uvula, base of the tongue, and nasopharyngeal airway (Fig. 2-15). Any sizable soft tissue pathology results in deviation or effacement of the airway. The tongue constitutes the bulk of the soft tissues at the level of the oropharynx. In children, and sometimes in adults, prominent lymphatic tissues such as adenoids and palatine tonsils may encroach on the nasopharyngeal and oral airways. Lingual tonsils are located at the base of the tongue above the valleculae, which are air-filled pouches between the tongue base and the free margin of the epiglottis.
Figure 2-15 Normal airway structures seen on a lateral cervical spine radiograph: 1, Hard palate; 2, soft palate and uvula; 3, air-filled vallecula; 4, epiglottis; 5, air-filled pyriform sinus; 6, air-filled stripe of laryngeal ventricle; C, noncalcified cricoid cartilage; h, hyoid bone; HP, hypopharynx; NP, nasopharynx; OP, oropharynx; t, thyroid cartilage.
The epiglottis is an elastic fibrocartilage shaped like a flattened teardrop or leaf that tapers inferiorly and attaches to the thyroid cartilage. The epiglottis tends to be more angular in infants than in adults. During the first several years of life, the larynx changes its position in the neck.23,24 The free edge of the epiglottis in neonates is found at or near the C1 level, and the cricoid cartilage, representing the most caudal portion of the larynx, is at the C4-C5 level. By adolescence, the epiglottis is found at the C2-C3 level and the cricoid is at the C6 level. The adult epiglottis is usually seen at the C3 level, with the cricoid at C6-C7. However, the position of these structures in the normal population varies by at least one vertebral body level.
Sometimes visualized by a cervical spine radiographic study with soft tissue neck technique or on the CT scout view or the MR sagittal view of the neck is a transversely oriented, air-containing lucent stripe, located just below the base of the aryepiglottic folds, which indicates the position of the air-filled laryngeal ventricle (Fig. 2-16). This marks the position of the true vocal cords, which are just below this lucent stripe. Lateral to the aryepiglottic fold is the pyriform sinus of the pharynx. This anterior mucosal recess lies between the posterior third of the thyroid cartilage and the aryepiglottic fold. The extreme lower aspect of the pyriform sinus is situated between the mucosa-covered arytenoids and the mucosa-covered thyroid cartilage, at the level of the true vocal cords. The air column caudally represents the cervical trachea. On the AP view, the false and true vocal cords above and below the laryngeal ventricles may be identified, as well as the subglottic region and the trachea.
Figure 2-16 Normal airway structures seen on a computed tomographic lateral scout view (A) and on a T1-weighted, fat-suppressed postcontrast sagittal magnetic resonance cervical spine study (B): 1, Hard palate; 2, soft palate and uvula; 3, retropharyngeal or prevertebral soft tissue; 4, epiglottis; 5, arytenoid prominence; 6, trachea air column; h, hyoid bone; HP, hypopharynx; LV, laryngeal ventricle; NP, nasopharynx; OP, oropharynx.
The landmarks dorsal to the airway are shadows representing the normal soft tissue structures of the posterior wall of the nasopharynx, which is closely adherent to the anterior surface of the atlas and the axis and extends superiorly to the clivus and inferiorly to become continuous with the soft tissues of the posterior wall of the hypopharynx. The ligaments of the cervicocranium are critical to maintaining stability throughout this region; they are directly involved in the range of motion of the cervicocranium and anteriorly contribute to the prevertebral soft tissue shadow. Superimposed on these deep structures are the constrictor muscles and the mucosa of the posterior pharyngeal wall. The cervicocranial prevertebral soft tissue contour should normally be slightly posteriorly concave rostral to the anterior tubercle of C1, anteriorly convex in front of the anterior tubercle, and posteriorly concave caudal to the anterior tubercle, depending on the amount of adenoidal tissue and on the amount of air in the pharynx.
Adenoidal tissue appears as a homogeneous, smoothly lobulated mass of varying size and configuration. The anterior surface of the adenoid is demarcated by air anteriorly and inferiorly. The air inferior to the adenoids allows differentiation between adenoids and the presence of a nasopharyngeal hematoma, which is commonly associated with major midface fractures. In infants and young children, the soft tissues of the cervicocranium are lax and redundant. Depending on the phase of respiration and position, the thickness of the prevertebral soft tissues may appear to increase and may simulate a retropharyngeal hematoma. This finding may extend to the lower cervical spine. This anomaly becomes normal if imaging is repeated with the neck extended and during inspiration. By 8 years of age, the contour of the soft tissues should resemble that seen in adults. Of note, in pediatric patients, sedation may result in a decrease in AP diameter of the pharynx at the level of the palatine tonsils, in the soft palate, and at the level of the epiglottis.
In the lower neck (C3 to C7), the prevertebral soft tissue shadow differs from that in the cervicocranium because of the presence of the beginning of the esophagus and the prevertebral fascial space, which are recognized on the lateral radiograph as a fat stripe. By standard anatomic description, the esophagus begins at the level of C4; however, in vivo, the esophageal ostium may normally be found as high as C3 or as low as C6 and varies with the phase of swallowing and the flexion and extension of the cervical spine.25 The prevertebral soft tissue thickness, the distance between the posterior pharyngeal air column and the anterior portion of the third or fourth vertebra, should not exceed one half to three quarters of the diameter of the vertebral body. In the opinion of Harris and Mirvis, only the measurement at C3 is valid, and it should not exceed 4 mm (Fig. 2-17).12
Figure 2-17 Prevertebral soft tissues seen on lateral cervical spine radiographic studies. A, Normal adult. B, Retropharyngeal abscess in a child. The arrow points to the anteriorly displace airway and increase in the prevertebral space between the vertebral column and the airway.
(Courtesy of Dr. Alan Schlesinger, Texas Children’s Hospital, Houston, TX.)
More caudally, at the cervicothoracic junction, assessment of the prevertebral soft tissues is based on contour rather than actual measurement. This contour should parallel the arch formed by the anterior cortices of the lower cervical and upper thoracic vertebral bodies.
In truth, plain film diagnosis of upper airway diseases has been supplanted by cross-sectional imaging, except in a few situations in which plain radiographic findings are pathognomonic of the disease. Two classic examples of plain film radiologic diagnosis are acute epiglottitis and croup.
In acute epiglottitis (or supraglottitis, a more encompassing term), there is edema and swelling of the epiglottis with or without involvement of the aryepiglottic folds and arytenoids. The offending organism is usually Haemophilus influenzae. Airway compromise with a rapidly progressive course requiring emergency tracheostomy is a possibility if the entity goes unrecognized and untreated. In general, the infection is milder in adults than in children. This entity, usually a more indolent form, is making a comeback among patients with acquired immunodeficiency syndrome (AIDS).
The findings on plain film are swelling or enlargement of the epiglottis. On the conventional lateral radiograph of the neck, thickening of the free edge of the epiglottis can be appreciated and is referred to as the “thumb sign” (Fig. 2-18). The width of the adult epiglottis should be less than one third of the AP width of the C4 body. Cross-sectional imaging is superfluous. However, the degree of airway compromise can theoretically be quantified by 3-D reformation.
In laryngotracheobronchitis or croup, the subglottic larynx is involved. This condition affects younger children and has a less fulminant course than acute epiglottitis. The swelling of the soft tissues in the subglottic neck can be appreciated on an AP view of the neck (Fig. 2-19). There is usually a long segment narrowing of the glottis and subglottic airway with loss of the normal angle between the vocal cords and the subglottic airway. This has been referred to as the “steeple sign.” The hypopharynx is usually dilated because of the airway obstruction distally.
Figure 2-19 Croup. Anteroposterior soft tissue neck radiograph in an infant. Long segment narrowing of the subglottic airway is present (arrow) with loss of the normal angle between the vocal cords and the subglottic airway (“steeple sign”).
(Courtesy of Dr. Alan Schlesinger, Texas Children Hospital, Houston.)
Plain films are usually obtained in the initial assessment of suspected foreign body ingestion. In children, up to 50% of witnessed foreign body ingestions are asymptomatic.26 Most foreign bodies are radiopaque, but wood and plastic usually are not visible on plain films. The radiopacity of ingested fishbone varies with the type of fish.27 In the neck, ingested foreign bodies most often lodge at the level of the pyriform sinus (Fig. 2-20).
Before the advent of CT, chest radiography was routinely ordered to assess pulmonary and cardiovascular status, and it is still a cost-efficient examination that yields a great deal of general information. The most common views of the chest are the posteroanterior (PA), anteroposterior (AP), and lateral projections (Fig. 2-21). The PA chest view is obtained with the patient’s anterior chest closest to the film cassette and the x-ray beam directed from a posterior to an anterior direction. Alternatively, the AP chest view is done with the patient’s back closest to the film cassette and the x-ray beam directed in the anterior to posterior direction. The part of the chest closest to the film cassette is the least magnified; therefore, the cardiac silhouette is larger on the AP projection. The lateral projection is most often performed with the patient’s left chest closest to the film cassette for better delineation of the structures in the left hemithorax, which is more obscured by the heart on a PA projection.
Normal posteroanterior (PA) and lateral radiographic studies of the chest. A, Normal PA view of a woman with increased density at the lung bases related to overlying breast tissues. B, PA view of a man with lucent lungs. C, Normal lateral chest view. D, Lateral view of a patient with chronic obstructive pulmonary disease showing barrel-shaped chest with increased retrosternal air; notice that the lung base appears progressively more lucent overlying the dorsal spine. The carina is indicated by an asterisk. 10, 10th posterior rib; A, aorta; Lt, left bronchus; Rt, right bronchus; T, trachea.
Other common projections include the oblique, decubitus, and lordotic views. The oblique view is useful for assessing a lesion with respect to other structures in the chest. The decubitus view is helpful to assess whether an apparent elevated hemidiaphragm is being caused by a large subpulmonic pleural effusion. The lordotic view is helpful to look for a suspected small apical pneumothorax, which can also be accentuated on an expiratory-phase examination.
It is useful to train one’s eyes to analyze the chest radiograph systematically to cover the details of the chest wall, including the ribs, lungs (field and expansion), and mediastinal structures such as the heart and the outline of the tracheal-bronchial tree. On an adequate inspiratory film, the hemidiaphragms are below the anterior end of the sixth rib, or at least below the 10th posterior rib, and the lung expansion should be symmetrical. The right hemidiaphragm is usually half an interspace higher than the left, which is depressed by the heart (see Fig. 2-21A). Without doubt, the art of chest radiograph interpretation has diminished since the advent of CT, which demonstrates chest pathology with unparalleled clarity. However, chest radiography can still provide a composite survey of the chest at one quick glance. One can easily compare the lung volumes, identify the position of the mediastinum, determine the presence or absence of major airspace disease, and make a gross assessment of the cardiac status.
A high hemidiaphragm implies reduced lung volume, which can result from phrenic nerve paralysis, thoracic conditions causing chest pain that leads to splinting, or extrapulmonic processes such as an enlarged spleen or liver, pancreatitis, or subphrenic abscess. The presumed level of the hemidiaphragm is seen as an edge or transition between the aerated lungs and the opacity of the organs in the abdomen. If the thin leaves of the hemidiaphragm are outlined by air, a pneumoperitoneum should be considered (Fig. 2-22).
Figure 2-22 Pneumoperitoneum. Postoperative anteroposterior chest radiograph after thoracotomy. Arrows in right upper chest demarcate the thin pleural line defining a tiny pneumothorax. Arrow in right lower chest indicates the right hemidiaphragm outlined by a small pneumoperitoneum. This patient also has cardiomegaly and right midlung and left basilar atelectasis.
A well-expanded lung should appear radiographically lucent but be traversed by “lung markings,” thin threads of interstitium consisting of septa and arterial, venous, and lymphatic vessels. In most normal individuals, the lungs appear more lucent at the top owing to the distribution of the pulmonary vasculature, the effect of gravity, and overlying soft tissues such as breast tissues. In patients with congestive heart failure or pulmonary venous hypertension, this pattern is reversed, with “cephalization” and engorgement of the pulmonary veins in the upper lung zones (Fig. 2-23; also see Fig. 2-21D). In general, any process such as fluid, pus, or cells that replaces the airspaces of the lungs causes the x-ray beam to be more attenuated, allowing less of the beam to be transmitted through the patient to the film. This causes the affected areas to appear less dark or more opaque (white) on the film. A whole host of diseases could be responsible, depending on the clinical picture, including pleural effusion, pulmonary edema, pneumonia, lung mass, lung collapse or atelectasis, lung infarct or contusion, and metastatic disease (Fig. 2-24). The key from an anesthesiologist’s point of view is not to make the correct pathologic diagnosis but to note the abnormality, which may affect ventilation, and adjust the anesthetic practice accordingly.
Figure 2-24 A and B, Left pleural effusion. Posteroanterior (PA) view of the chest (A) shows almost complete “white-out” of the left hemithorax and minimal residual aerated left upper lung zone. There is a mass effect with deviation of the trachea to the right. On the lateral view (B), the pleural effusion is less apparent. The tipoff is the lack of the expected lucency overlying the spine at the base (compare Fig. 2-21, C and D). C, Pulmonary edema. Anteroposterior (AP) view of the chest demonstrates bilateral hazy lung fields with air bronchogram. A tracheostomy tube is present. D through F, Left lower lung mass. Notice that although the inspiratory effort is the same on both the PA (D) and the AP (E) view (i.e., hemidiaphragm below ninth posterior rib), the cardiac silhouette and the left lower lobe mass appear larger on the AP view by virtue of the film geometry and magnification factor. The lateral view (F) helps to localize the disease process to the lateral segment of the left lower lobe. A mass is noted with postobstructive atelectasis (arrows). G and H, Aspergillosis. AP chest radiograph (G) shows nodular densities in both lungs. The differential diagnosis includes inflammatory and neoplastic processes. Notice that the tip of the endotracheal tube is in a good position, above the carina, and there is a central line on the right. Axial computed tomogram of the chest (H) better demonstrates the nodular pattern of lung involvement. I and J, Melanoma metastases to the lungs. The PA (I) and lateral (J)