7 Imaging of Oral Cavity Cancers
Radiologic imaging plays an important role in the assessment of oral cavity cancers. CT and MRI are complementary in evaluation of cancers in the subsites of the oral cavity and their routes of spread. Important features to assess on CT and MRI include but are not limited to the depth of invasion, extension across the midline, invasion of the neurovascular bundle, perineural spread, osseous invasion, and nodal metastasis.
The oral cavity is the most ventral portion of the aerodigestive tract and has a complex anatomy with multiple anatomic subdivisions. The separation between the oral cavity and the oropharynx is very important. The squamous epithelium of the oral cavity derives from ectoderm, which tends to give rise to more differentiated and less aggressive lesions. In contrast, the squamous epithelium of the oropharynx derives from endoderm, which tends to give rise to poorly differentiated and aggressive carcinomas. 1
The oral cavity is separated from the oropharynx by a ring of structures consisting of the circumvallate papillae, the hard and soft palate junction, and the anterior tonsillar pillars on both sides. 1 Please refer to the section on “Diagnosis and Evaluation” for a detailed description of the oral cavity anatomy and its subdivisions. The most common sites of involvement are the lower lip, lateral and ventral surface of the oral tongue, and the floor of the mouth. 1 , 2
Cancer of the oral cavity accounts for nearly 30% of all head and neck malignancies, and approximately 300,000 new cases of oral cavity cancers are diagnosed in the world every year. 3 The vast majority (almost 90%) of oral cavity malignancies are squamous cell carcinomas (SCCs), and the rest are comprised of minor salivary gland neoplasms, melanoma, lymphoma, and rare varieties of squamous cell and odontogenic tumors. 2 , 4
Oral cavity cancer occurs most frequently in individuals older than 50 years of age, affecting men more frequently than women with a ratio of 2:1 to 3:1. In the United States, there is a higher incidence of oral cancer in the African American population. The main risk factors for the development of oral cavity cancer are tobacco and alcohol use. 2 , 5 , 6
In the developed world, the overall incidence of oral cavity cancer has been declining since the 1970s due to the decrease in prevalence of tobacco use. 7 , 8 However, the prevalence of tobacco use is rising in the rest of the world, and the future burden of oral cavity cancer is expected to significantly increase in the lower-income countries, with an expected 80% increase in oral cavity cancers. 9
Tobacco use is the most important risk factor for oral cavity cancer, and approximately 65% of oral cavity cancers are attributed to smoking tobacco. Tobacco use, including both smoking and chewing, confers a threefold to fivefold increased risk of oral cavity cancer, which is independent of alcohol use. However, the synergistic effect of alcohol and tobacco on increasing the risk of oral cavity cancer is well established, and results in almost threefold greater risk than the product of the individual effects of tobacco and alcohol. 6
Additional risk factors include chewing betel quid leaves and areca nuts, which contribute to the high incidence of oral cavity cancers in Central and Southeast Asia. Prolonged ultraviolet (UV) exposure in regions with hot climates increases the risk of lip cancer. 10 Increased incidence of lip cancer has also been shown in renal transplant recipients on immunosuppressive therapy. 11
Human papillomavirus (HPV) is now established as a major risk factor for oropharyngeal cancer in the developed world, but its role in oral cavity cancer is very limited. Multiple studies have shown that HPV does not often play a role in the pathogenesis of oral cavity cancer. In the United States, high-risk HPV type 16 deoxyribonucleic acid (DNA) has been detected in only 3 to 10% of oral cavity cancers, whereas it has been detected in nearly 60 to 80% of oropharyngeal cancers. 12 , 13 The difference in HPV prevalence in the oral cavity cancers and oropharyngeal cancers has implications for prevention, diagnosis, workup, treatment, and follow-up. 14
Fanconi’s anemia is a rare disorder that confers a 500- to 700-fold increased risk of developing head and neck SCC, most of which occur in the oral cavity. In patients who present with head and neck cancer that are younger and without typical risk factors of tobacco and alcohol use, Fanconi’s anemia should be considered. Fanconi’s anemia has an incidence of approximately 1:360,000 births, but the carrier frequency is believed to be as high as 1:100 in certain populations, such as Ashkenazi Jews, Spanish gypsies, and black South Africans. It is inherited in an autosomal recessive fashion, and it is the most common form of inherited aplastic anemia. 15
7.3 Clinical Presentation
Oral cavity cancer is typically painless and asymptomatic at an early stage. As a result, patients with readily visible lip cancers present at an early stage. However, patients with intraoral cancers usually present when they are symptomatic and at a more advanced stage. Most common presenting symptoms are painful oral lesion, painful or bleeding ulcer, and loose teeth or ill-fitting dentures. Patients may also present with trismus (caused by pterygoid involvement), hypoesthesia (caused by perineural involvement), or otalgia (referred pain from cranial nerves [CNs] IX and X). 2 , 16 , 17
On examination, tumors may appear as white, red, or speckled mucosal lesions, and advanced lesions classically have an ulcerated center with indurated or indistinct borders. 2 The most common locations for intraoral lesions are the lateral and ventral aspects of the oral tongue, and the floor of the mouth. 18
A substantial percentage of patients have cervical lymphadenopathy at the time of initial presentation. Careful physical and radiologic evaluation of the cervical lymph node chains is important as the presence or absence of lymphadenopathy is the most important prognostic indicator in this disease. 17 , 19
7.4 Diagnosis and Evaluation
Radiologic imaging plays an important role in evaluation and staging of oral cavity cancer. CT, MRI, and PET/CT are important in assessment of tumor size and depth of invasion (DOI), bone invasion, lymphadenopathy, presence of nodal extracapsular extension, perineural spread, second synchronous lesions, and distant metastases. In the section, we will present an overview of imaging modalities and patterns of spread for oral cavity cancers.
7.4.1 Computed Tomography
Contrast-enhanced CT is the most commonly utilized imaging modality for initial assessment of oral cavity cancer. CT is readily available and can be acquired very quickly, which limits motion artifact, and can be easily reformatted into coronal and sagittal planes. CT provides exquisite evaluation of cortical bone involvement. The disadvantages of CT include exposure to ionizing radiation and potential adverse reactions related to iodinated contrast agents, as well as streak artifact from dental amalgam and the mandible (▶ Fig. 7.1a). However, scanning in multiple planes can overcome the limitations related to streak artifacts. 2
On CT, SCCs appear as homogeneous or heterogeneous soft-tissue attenuation lesions that demonstrate variable contrast enhancement. Early or more prominent enhancement of the tumor margins may be present, which is presumably related to higher vascularity at the periphery of the tumor. Necrotic areas demonstrate central hypodensity with peripheral enhancement. The peripheral enhancement can sometimes have similar attenuation as surrounding soft tissues, which can limit evaluation of the exact boundaries of the lesion. 2 On a technical note, it is important to perform the scan during optimum contrast enhancement of tissues. With the very fast modern scanners, it is possible to scan too early before the peak enhancement of the tumor occurs. In this regard, the timing of scanning with respect to contrast material injection is different in tumor assessment from other neck CT applications such as CT angiogram.
7.4.2 Magnetic Resonance Imaging
MR imaging offers better soft-tissue contrast and can complement CT in accurately delineating tumor extent and thickness. On conventional non-fat saturated T1-weighted images, SCC has intermediate signal intensity and generally demonstrates iso- to hypointensity compared to muscles of the floor of the mouth and tongue. Even though tumors have similar intensity to muscle on T1-weighted images, this sequence can still be very helpful to determine tumor margins as its boundaries may still be evident through disrupted internal structure of normal soft tissues. On T2-weighted images, SCC is typically heterogeneously hyperintense to normal muscle, making it more conspicuous, although it may be difficult to distinguish tumor from adjacent surrounding edema and inflammation resulting in overestimation of tumor size. On STIR (short tau inversion recovery) images, tumors have similar appearance as on T2-weighted images and may be more conspicuous, but STIR images are more prone to artifact and usually have lower spatial resolution. It should be noted that normal mucosal surfaces may have high signal intensity on STIR images. 2
On gadolinium-enhanced T1-weighted images without fat saturation, SCC demonstrates variable enhancement and conspicuity (▶ Fig. 7.1c). Fat-saturated images can be helpful in distinguishing tumors from normal surrounding intrinsically hyperintense structures. However, some normal structures, such as salivary glands, also demonstrate increased conspicuity on fat-saturated images, and should not be mistaken for pathology. Fat-saturated images require longer acquisition time, which may result in increased motion artifact. Susceptibility artifacts near bone or air limit evaluation of skull base foramina on the fat-saturated images.
MRI is superior to CT for assessment of perineural spread of disease, and medullary bone involvement (see below). MRI also has the advantage of reduced artifact from dental amalgam. However, MRI is more affected by motion artifact. 2
7.4.3 Fluorodeoxyglucose PET/CT
Fluorodeoxyglucose (FDG) PET/CT has the potential to increase sensitivity for nodal disease, but studies have shown that it is generally not substantially superior to standard contrast-enhanced CT or MR imaging in detecting occult nodal metastases. 16 , 20 – 22 FDG-PET/CT is often utilized to detect an occult primary head and neck malignancy in the setting of a positive cervical lymph node. Finally, FDG-PET/CT is also useful in detecting second primary lesions and distant metastases. However, the positive predictive value for distant metastases is low, and often results in unnecessary additional imaging and biopsies. Therefore, FDG-PET/CT should be used for detection of distant metastases in the setting of advanced local disease with higher risk of distant metastases. 23 , 24
7.4.4 Nodal Metastasis
The imaging assessment of cervical lymphadenopathy must include a determination of size, morphologic features, and margination of the node. The cardinal features of pathologic lymphadenopathy on imaging are large node size and presence of central necrosis. The usual size criterion as a maximal longitudinal diameter is 15 mm or greater for jugulodigastric nodes, and 10 mm or greater for other nodes. The usual size criterion as a minimal axial diameter is 11 mm for jugulodigastric nodes and 10 mm for other nodes. Clustering of three or more nonenlarged lymph nodes in the same lymphatic drainage area is considered suspicious for pathologic lymphadenopathy. A rounded appearance with loss of the normal reniform shape of a lymph node is also suspicious for nodal metastasis. 1 , 2 , 17
Extracapsular spread (ECS) is an important finding as it results in an approximately 50% reduction in survival. 16 , 25 Imaging features of extracapsular spread on CT includes presence of an irregular, spiculated, blurred lymph node border; loss of fat plane around the node; irregular capsular enhancement; central nodal necrosis; and/or tumoral infiltration into soft tissue and vascular structures (▶ Fig. 7.2, ▶ Fig. 7.3, ▶ Fig. 7.4). On MR imaging, the ranked predictive criteria for extracapsular spread include high signal intensity in tissues surrounding a node on T2 STIR images, a poorly defined nodal border, irregular rimlike enhancement, and large nodal size. 1 , 26
7.4.5 Perineural Spread
Perineural spread makes local and regional control of the disease particularly difficult by allowing tumor extension beyond the expected treatment margins. It is important for the radiologist to be alert to the possibility of perineural spread as it is often clinically silent. Perineural invasion is considered a characteristic of aggressive SCCs, and it is more common in subsites such as the floor of the mouth and hard palate, where the neurovascular bundles are particularly accessible. 1
On CT, findings associated with perineural spread include foraminal enlargement and soft-tissue attenuation replacing fat attenuation within the foramina. MRI is superior to CT for direct assessment of perineural spread. Findings that are suspicious for perineural spread on MRI include enlargement, irregularity, and abnormal enhancement of the major cranial nerves and their branches. Replacement of high-intensity fat adjacent to the major skull base foramina and other fat-containing spaces such as the pterygopalatine fossa (PPF) by intermediate signal intensity tissue on unenhanced T1-weighted images are also suspicious findings. Indirect signs of nerve invasion such as muscle denervation may be seen as well. In the acute phase, a denervated muscle may appear swollen and hyperintense on T2-weighted images with increased enhancement. In the chronic phase, muscle denervation results in muscle atrophy and fatty replacement. 2
7.4.6 Mandibular Invasion
Osseous invasion of oral cavity cancers upstages the disease to at least stage T4a, and presence or absence as well as the degree of osseous invasion alters treatment strategy. Tumors of the floor of the mouth, ventral surface of the tongue, gingiva, and retromolar trigone (RMT) may show early involvement of the mandible. Mechanism of invasion and routes of entry into the mandible have been studied, and the current understanding is that the tumor invades the mandible by direct extension through the periosteum rather than through lymphatic spread. 27 In the dentate mandible, tumors of the floor of the mouth and ventral surface of the tongue enter the mandible most commonly at the point of abutment rather than through the occlusal surface, neural foramina, or the periodontal membrane. The point of tumor abutment is commonly at the junction of the attached and reflected mucosa, and therefore, small cortical defects at the attachment of the mucosa to the bone are likely the weak points through which the tumor gains entry into the mandible. 28
In the edentulous mandible, cortical irregularities in the occlusal surface of the mandible may serve as entry points for oral cancers into the mandible. Tumors of the gingiva and the RMT may also preferentially enter the mandible through the occlusal surface, which is commonly the point of abutment of these tumors with the mandible. 28 – 30
Tumors invading the mandible can be managed with either a marginal or a segmental resection. If there is invasion of only the periosteum without gross cortical invasion, the tumor can be managed with a marginal mandibulectomy maintaining the integrity of the mandible. In general, tumors with gross cortical erosion or invasion of the medullary cavity require a segmental resection, which involves resection of the entire involved segment of the mandible necessitating a more complex reconstruction, requiring vascularized bone and soft tissue, often a fibular free flap. 17
CT and MRI are complementary for evaluation of mandibular involvement. CT is superior to MRI for evaluation of subtle cortical erosion with high specificity and positive predictive value (90%), but limited sensitivity (50%; ▶ Fig. 7.5). 31 A pitfall in the evaluation of cortical erosion can occur in the body of the mandible when the plane of imaging is not along the long axis of the mandibular body (▶ Fig. 7.6).
MR imaging is superior to CT for evaluation of the medullary cavity with high sensitivity (▶ Fig. 7.7), but low specificity. 32 Signs of marrow invasion on MR include T1 hypointensity replacing normal fatty marrow, loss of low-signal-intensity cortex, and contrast enhancement within the bone or along the inferior alveolar nerve. However, MR imaging suffers from low specificity in the setting of odontogenic disease, which can cause false-positive marrow replacement and even enhancement. 17 , 33
Staging of oral cavity SCC and minor salivary gland carcinomas is based on the tumor, node, and metastasis (TNM) staging system, which has been revised in the eighth edition of the American Joint Committee on Cancer (AJCC) Cancer Staging Manual (▶ Table 7.1) with significant changes compared to the seventh edition. 34 Recent data have shown that DOI is a better prognostic indicator than tumor thickness for oral cavity cancers. 35 Pathologically, DOI is measured from the level of the basement membrane of the closest adjacent normal mucosa, and a “plumb line” is dropped from this plane to the deepest point of tumor invasion. DOI will now affect the T stage, which will increase by one level for every 5 mm increase in DOI. DOI supersedes not only tumor thickness, but also extrinsic tongue muscle involvement, which is no longer part of the criterion for T4 designation.
Patterns of invasion determine T4a or T4b stage of the lesion. Osseous invasion is the primary consideration for upstaging to T4a disease, or “moderately advanced” disease. T4b disease, or “very advanced” disease, refers to tumor invasion of the masticator space, pterygoid plates, or skull base; or tumor encasement of the internal carotid artery.
The criteria for nodal staging for oral cavity cancers are identical to the criteria for many other sites in the head and neck such as the pharynx and larynx. A significant change has been made to the N category of the staging system for head and neck cancers in the AJCC eight edition with inclusion of extranodal extension (ENE) as a prognostic variable for lymph node metastases. There is long-standing evidence that ENE profoundly affects prognosis for oral cavity cancers. 36 Current imaging modalities lack sensitivity and specificity in their abilities to detect early or minor ENE, and imaging evidence alone may be supportive but is not sufficient. 37 Therefore, only unambiguous ENE will be assigned a status of ENE positive, as determined by physical examination (invasion of skin, infiltration of musculature/dense tethering to adjacent structures, or dysfunction of a cranial nerve, the brachial plexus, the sympathetic trunk, or the phrenic nerve) and supported by radiological evidence. 34