Ryan T. Fitzgerald Abstract
Cross-sectional imaging has become an integral part of the workup, treatment planning, and surveillance for patients with neoplasms arising within or otherwise involving the head and neck and directly impacts decision-making regarding neck dissections. This chapter is organized around imaging anatomy as it applies to neck dissection, imaging approaches to staging, and surveillance imaging. Much of the discussion revolves around in computed tomography (CT), but MRI and combined CT/PET strategies are also briefly touched upon.
Keywords: CT, MRI, neck dissection, lymphadenopathy, head and neck cancer, squamous cell carcinoma
13.1 Introduction
Imaging plays an integral role in the staging, treatment planning, and surveillance for squamous cell carcinoma (SCC) and other malignancies occurring in the head and neck. Advances in CT and MRI in recent years have led to improvement in each of these imaging modalities as they apply to the management of head and neck cancer including their role in determination of the necessity of neck dissection.
Modern imaging and its interpretation are challenging endeavors due to the complexity of the hardware and software technologies that form the basis of image generation. Although this chapter will cover some important technical considerations that impact image acquisition and interpretation, the bulk of the material will focus on how imaging can be applied to maximize diagnostic certainty while at the same time maintaining efficiency and cost-effectiveness. Discussion of neck anatomy as it applies to the search for, and description of, cervical lymph node disease is covered briefly herein and more extensively in other chapters. Rather than organizing this chapter around imaging modalities, an outline based on imaging features of nodal metastases across multiple imaging modalities will instead be followed. Given its high prevalence relative to other malignancies, the bulk of the text is based on data from studies of nodal disease attributable to SCC. Nevertheless, many of the concepts such as imaging features indicative of lymph node metastases are broadly applicable across a wide spectrum of neoplasms. In the face of ever-evolving imaging techniques applied to head and neck malignancy, the goal has been to collate peer-reviewed evidence from the past decade. That said, in some cases it has been necessary to cite earlier work upon which more recent investigations have built. After a short synopsis of imaging-based neck anatomy, the bulk of this chapter will focus on the imaging approach to neck staging and thereafter conclude with a discussion of the imaging approach to surveillance of the neck.
13.2 Imaging Anatomy
Image-based classification of cervical lymph nodes in the context of head and neck malignancy has been widely adopted due to the ubiquity of imaging as part of cancer staging for most patients, the ability of imaging to detect clinically occult nodal metastases, and the high level of reproducibility with which modern cross-sectional imaging can localize metastatic nodes in relation to anatomic landmarks.1 In 2000, Som et al proposed the most widely utilized radiologic classification system for cross-sectional imaging assessment of cervical lymph nodes, which complements the clinically based classification espoused by the American Joint Committee on Cancer and the American Academy of Otolaryngology-Head and Neck Surgery.1 In this system, landmarks forming the borders of each nodal level are based on axial image sections. Level I includes lymph nodes superior to the hyoid bone, below the mylohyoid muscle, and anterior to the posterior borders of the submandibular glands. Levels II and III refer to lymph nodes that reside anterior to a transverse line along the posterior borders of the sternocleidomastoid muscles. Level II nodes are posterior to the posterior margin of the submandibular gland, lie between the skull base and inferior margin of the hyoid bone, and are positioned lateral to the medial border of the internal carotid artery. Below level II, level III describes a contiguous compartment that extends inferiorly to the lower border of the cricoid cartilage. Further inferiorly, level IV extends to the level of the clavicle. In contrast to levels II and III, level IV nodes lie anterior and medial to an oblique line drawn through the posterior edge of the sternocleidomastoid muscle and the posterior lateral edge of the anterior scalene muscle. Level V describes nodes that are dorsal to the posterior border of levels II to IV and lie anterior to a transverse line through the anterior edge of the trapezius muscles. Level VI refers to lymph nodes that lie inferior to the lower body of the hyoid bone, superior to the upper border of the manubrium, and medial to the medial borders of the common/internal carotid arteries. Level VII lymph nodes lie between the upper border of the manubrium and brachiocephalic veins and are also medial to the medial borders of the common carotid arteries. CT- based examples of the Som classification scheme are available in their original publication.1
Anatomic localization of suspected nodal metastases is important not only for staging and treatment planning, but can also provide guidance for further investigation if the site of primary neoplasm is not readily apparent. For example, nodal metastases isolated to level II would draw attention to the oral cavity or oropharyngeal mucosal space rather than the thyroid gland, whereas the finding of abnormal nodes confined to levels VI and VII would be much more likely to emanate from a thyroid primary. In cases in which the primary neoplasm is known, such knowledge can direct scrutiny to first-level nodal drainage areas and thus maximize the detection of early or subtle signs of metastatic involvement of cervical lymph nodes.
13.3 Imaging Approach to Neck Staging
In cases of SCC arising in the head and neck as well as other neck neoplasms, the presence or absence of metastatic lymph
nodes substantially impacts treatment options and prognosis. Given the accuracy of modern cross-sectional imaging to identify and localize nodal metastases, imaging has become a necessary adjunct to the clinical examination for treatment planning. CT is currently the most widely utilized imaging modality for neck staging owing to its wide availability and technical capabilities. Much of the following section will thus focus on the application of CT in the neck staging, largely due to the fact that the preponderance of literature on the subject is CT based. MRI may also play a role in select cases such as patient for whom io- dinated CT contrast agents are contraindicated (e.g., prior anaphylactic reaction after iodinated contrast) or for problem solving in the setting of suspected skull base invasion or perineural tumor spread owing to the superior sensitivity of MR for these applications. Among the barriers to expanded use of MRI are longer scan times relative to CT, increased cost, issues of claustrophobia, and the direct relationship between scan time and image degradation secondary to patient motion.
13.4 Lymph Node Characteristics Related to Metastatic Involvement
13.4.1 Size
Although size is frequently invoked in the discussion of lymph node metastases, it is a poor marker of metastatic involvement relative to other factors discussed later. At many centers, lymph node size is reported as long-axis diameter on an axial image section and may also include a long-axis measurement in another plane if such a measurement would impact staging. Long- axis measurements, in contrast to short-axis measurements that are employed at other locations such as the mediastinum, best replicate the clinical determination of palpable lymph node enlargement. Measurement in the long axis is the most frequently employed methodology across the literature and is the metric utilized in the TNM classification of the American Academy of Otolaryngology-Head and Neck Surgery.
Underlying the limitations of lymph node size as a primary determinant of metastatic involvement is the frequency of pathologic confirmation of metastases in lymph nodes of normal size (Fig. 13.1). As an example, in a study of subjects with head and neck SCC, Don et al found that 67% of metastatic nodes had a longitudinal diameter smaller than 1 cm.2 Lymph node enlargement may be a late manifestation of metastatic involvement that is proceeded by heterogeneous enhancement or other more sensitive imaging biomarkers.3 The specificity of lymph node size as an indicator of metastatic involvement is lacking due to the propensity of lymph nodes to become enlarged as a response to infection and inflammation. The existence of multiple disagreeing size criteria and differences in accepted size thresholds based on location and patient age further confounds the application of size to the staging assessment.3 For instance, the application of a 1cm threshold within levels II and VI would yield a substantial difference in the specificity of nodal metastases, as nodes measuring above 1 cm are frequently encountered in level II of normal subjects. As such, many practitioners use a tiered threshold in which the size threshold for levels I and II is higher than that for more caudal levels. Asymmetry can serve as another useful discriminator regarding the significance of lymph node enlargement, whereby enlarged lymph nodes that are unilateral and/or confined to a single anatomic level are more likely to be true positives than cases in which nodal enlargement is more widespread (Fig. 13.2). Despite the suboptimal sensitivity and specificity of size as a marker of nodal metastases, lymph node enlargement can be a useful tool toward identification of nodes requiring further scrutiny, but as an isolated feature it is not necessarily sufficient for the determination of metastatic involvement in the absence of additional markers of nodal metastases such as irregular margins or internal necrosis.
13.4.2 Morphology and Architecture
Physiologic lymph nodes are reniform (kidney) in shape and are composed of intermediate-density tissues surrounding an eccentric fatty hilum. As metastatic cells proliferate within a lymph node, concentric centrifugal expansion tends to alter the morphology of the node toward a more rounded shape. Thus, round lymph nodes, regardless of size, appropriately raise suspicion for metastatic involvement. In nonmetastatic reactive lymph nodes, the ratio of the long axis diameter to the short axis diameter exceeds 2:1 in 86% of cases.4
Lymph node enhancement is another feature that can be applied toward the determination of metastatic involvement. The parenchyma of physiologic lymph nodes enhances homogeneously, whereas nodes harboring metastases may show areas of heterogeneity, typically hypoenhancement, that is of moderate specificity for metastatic involvement. The finding of hyperphysiologic enhancement is both a poorly sensitive and poorly specific marker of metastatic involvement as it is commonly encountered in reactive lymph nodes (in the setting of infection/inflammation) and often persists in nodes previously treated by chemotherapy and radiation. Among the most specific markers of metastasis is the presence of intranodal necrosis (Fig. 13.3, Fig. 13.4). Careful attention may be required to distinguish between the physiologic fatty hilum and necrotic tissue, particularly in lymph nodes that are not significantly enlarged. CT has been shown to be superior to MR for the detection of nodal necrosis.5
Cystic nodal metastases can be encountered in metastatic SCC from any primary site; however, there is an established predilection for the development of cystic nodes in association within primary cancers arising within Waldeyer’s ring.6 Thus, the discovery of cystic lymph nodes, particularly across levels II and III, should prompt scrutiny of pharyngeal lymphoid tissue. More recently, an association has been described between cystic nodal morphology and human papillomavirus (HPV) positive SCC.78 In a study of 136 oropharyngeal SCC (OPSCC), metastases associated with HPV-positive tumors
demonstrated cystic features in 36% of cases versus in 9% of cases for nodal metastases attributed to HPV-negative pri- maries.8 Papillary carcinoma of the thyroid is another primary tumor that is frequently implicated in the development of cystic nodal metastases. Thus, cystic nodal metastases within levels IV, VI, and VII and sparing of level II/III would appropriately focus attention to the thyroid gland (Fig. 13.5).
13.4.3 Margins
Physiologic lymph nodes display clear, sharp margins bordered by homogeneous fat. Relative to size, the development of marginal irregularity is a specific marker of metastasis, as lymph nodes reacting to local of systemic infections may become enlarged but in most cases maintain normal margins. Extension of metastatic growth across and beyond the nodal capsule can also elicit stranding within adjacent fat. Loss of nodal margins should also prompt assessment for potential invasion of adjacent structures such as the carotid sheath vasculature, musculature, or glandular tissue, as such findings impact staging (Fig. 13.6). A discussion of lymph node margin status as it applies to extracapsular disease spread can be found in the following section.
13.4.4 Extracapsular Spread
Extracapsular spread (ECS) of nodal metastases from head and neck SCC has been established as a poor prognostic indicator in terms of 5-year overall survival and also confers an increased probability of locoregional recurrence and distant metastases.9.10 Although histologic determination of ECS is the gold standard, radiographically determined ECS also confers poor distant control of disease and attenuated survival.11 Pretherapy, radiologic determination of ECS status may hold particular importance for patients with oropharyngeal cancers (OPSCC), which are often HPV related and amenable to treatment with radiation alone for early-stage disease and combined radiation and chemotherapy in more advanced disease.12 Based on studies showing the survival benefit of chemotherapy in addition to routine postoperative radiation in OPSCC patients with positive margins or ECS, surgery could be reasonably deferred for OPSCC patients with radiologically determined ECS and reserved for cases that eventually require salvage therapy.13
While the potential for noninvasive determination of ECS status is promising, studies examining the test characteristics of CT and MRI for the detection of ECS have failed to show high degrees of reliability. A summary of reported sensitivities across studies using both CT and MR yielded suboptimal sensitivities ranging from 62.5 to 80.9% and specificities from 60 to 93%.13 A large study of 432 patients with head and neck SCC undergoing neck dissection found a CT-derived specificity of 97.7% although the sensitivity in the study was only 43%.14 Positive predictive values for CT-based ECS determination have been reported from 71 to 84% and negative predictive values from 48 to 49%.15 Given the impact of ECS determination on recent management trends, examination of techniques that could potentially boost the sensitivity of ECS detection will be important going forward.
Inclusion of both macroscopic and microscopic extracapsular spread of disease on histopathology as ECS positive is almost certainly responsible in part for the suboptimal sensitivity of imaging-based determination of ECS status based on the constraints of spatial resolution inherent in current techniques. The predilection of some head and neck cancer metastases, particularly those from HPV-positive OPSCC, to exhibit cystic features may also complicate the assessment for ECS. In their study of 111 patients with OPSCC, Aiken et al found that intranodal necrosis was the most robust radiologic predictor of pathologically proven ECS, whereas irregular borders and gross invasion approached, but did not meet, statistical significance.13 The overlap between the imaging appearance of internal necrosis and cyst formation may partially explain the lack of robust sensitivity of radiologic ECS determination across prior conventional cross-sectional imaging studies.
13.5 Imaging Approach to Surveillance
Imaging not only is applicable for staging and treatment planning but also plays an important role in the posttreatment setting for patients with head and neck malignancies in order to assess the response to treatment, guide further treatment, and also to establish a new baseline examination to which future radiologic surveillance studies can be compared. The timing and methods employed for surveillance vary across institutions, but in general patients who have received definitive radiation therapy and/or chemoradiotherapy undergo imaging 12 weeks after the conclusion of treatment. The typical 3-month gap between completion of therapy and imaging facilitates the evolution of treatment effects including tumor involution and thus reduces the possibility of false-positive results that can occur if imaging is obtained during or soon after treatment. For patients amenable to surgical treatment with curative intent, surveillance imaging is typically performed within the first 6 months after treatment.
CT continues to serve as the most widely utilized modality for imaging surveillance in patients with SCC of the head and neck for both the initial posttreatment examination and subsequent surveillance studies. The frequency and duration of surveillance radiographic assessment is patient specific and may vary according to pretreatment disease severity and treatment response. In general, surveillance imaging is performed every 3 to 12 months for up to 3 years. Any clinical suspicion of neoplastic recurrence during this time frame and beyond would appropriately prompt imaging for confirmation (ultrasound- or CT-directed biopsy) and restaging.
Surveillance imaging can place a burden on patients with head and neck cancer in terms of both cost and frequency. Rather than following a traditional CT-based algorithm, some groups have explored the application of PET/CT-based strategy. McDermott et al reported in 2013 that two consecutive negative PET/CT examinations within a 6-month period after the conclusion of treatment resulted in a negative predictive value for neoplastic recurrence of 98%, which could obviate the necessity of further radiologic imaging in the absence of clinical signs of recurrence.16 The potential of two PET/CT scans rather than a series of CT examinations over 3 years would likely result in an overall cost savings despite the higher relative cost of PET over CT. In a subset of head and neck cancer patients with advanced nodal disease (N2 or N3) receiving primary chemoradiotherapy, Mehanna et al showed that surveillance using combined CT/PET at 12 weeks, with neck dissection being performed only in cases showing an incomplete or equivocal response, was statistically equivalent to planned neck dissection in terms of survival.17 Further, patients randomized to the imaging arm not only underwent fewer surgeries, but also incurred a substantially reduced cost of treatment.17
Imaging features of metastatic nodal involvement in the pretreatment neck such as altered nodal morphology/margins and lymph node enlargement continue to be applicable following surgery, radiation, chemotherapy, and combined treatments. That said, disturbance of anatomic relationships due to surgery and radiation-induced changes within superficial and deep soft-tissue compartments often complicate the typical anatomy-based search pattern in the posttreatment neck. Nodal contrast enhancement is of particularly poor specificity in the posttreatment setting and can be attributed to therapeutic irradiation or various chemotherapeutic and immunomodulatory agents18 (Fig. 13.7, Fig. 13.8).
13.6 Conclusion
Imaging plays an integral role in the determination of the necessity for and/or extent of neck dissection/radiation. This chapter has outlined relevant imaging anatomy and techniques, discussed criteria for the determination of metastatic lymph node involvement, and explored the implications of imaging features of nodal disease. Although modern techniques have greatly advanced the value of cross-sectional imaging for staging and surveillance of neoplastic disease in the head and neck, it is nonetheless important to recognize their limitations. The sensitivity of CT and MRI is constrained by the fact that up to 25% of clinical N0 necks harbor micrometa- stases that are beyond the spatial resolution of even the best current cross-sectional techniques.19 Ultrasound and PET, covered in separate chapters, serve as complementary and/or adjunct modalities that further contribute to the workup, treatment planning, and surveillance for patients with head and neck cancers. Other burgeoning techniques such as molecular imaging are beyond the scope of the current text but may contribute to decision-making surrounding neck dissections in the future.
Fig. 13.7 Coronal (a) and axial (b) contrast- enhanced CT images show multiple enlarged, enhancing lymph nodes (arrows) bilaterally in this patient with a history of melanoma being treated with ipilimumab. See Fig. 13.8 for PET images and further discussion.
Fig. 13.8 Fluorodeoxyglucose (FDG) PET images from the same patient in Fig. 13.7 at baseline (a) and 2 months later (b) after the discontinuation of ipilimumab. The baseline examination (a) shows widespread FDG-avid lymphadenop- athy that had fully resolved at 8 weeks (b). Lymph node biopsy shortly after the baseline examination yielded reactive lymphadenitis and no malignant cells.
References
[1] Som PM, Curtin HD, Mancuso AA. Imaging-based nodal classification for evaluation of neck metastatic adenopathy. AJR Am J Roentgenol. 2000; 174(3):837-844
[2] Don DM, Anzai Y, Lufkin RB, Fu YS, Calcaterra TC. Evaluation of cervical lymph node metastases in squamous cell carcinoma of the head and neck. Laryngoscope. 1995; 105(7, pt 1):669-674
[3] Castelijns JA, van den Brekel MW. Imaging of lymphadenopathy in the neck. Eur Radiol. 2002; 12(4):727-738
[4] Bruneton JN, Balu-Maestro C, Marcy PY, Melia P, Mourou MY. Very high frequency (13 MHz) ultrasonographic examination of the normal neck: detection of normal lymph nodes and thyroid nodules. J Ultrasound Med. 1994; 13(2):87-90
[5] Curtin HD, Ishwaran H, Mancuso AA, Dalley RW, Caudry DJ, McNeil BJ. Comparison of CT and MR imaging in staging of neck metastases. Radiology. 1998; 207(1):123-130
[6] Goldenberg D, Sciubba J, Koch WM. Cystic metastasis from head and neck squamous cell cancer: a distinct disease variant? Head Neck. 2006; 28(7):633-638
[7] Goldenberg D, Begum S, Westra WH, et al. Cystic lymph node metastasis in patients with head and neck cancer: an HPV-associated phenomenon. Head Neck. 2008; 30(7):898-903
[8 ] Cantrell SC, Peck BW, Li G, Wei Q Sturgis EM, Ginsberg LE. Differences in imaging characteristics of HPV-positive and HPV-negative oropharyngeal cancers: a blinded matched-pair analysis. AJNR Am J Neuroradiol. 2013; 34(10):2005-2009
[9] Kokemueller H, Rana M, Rublack J, et al. The Hannover experience: surgical treatment of tongue cancer: a clinical retrospective evaluation over a 30 years period. Head Neck Oncol. 2011; 3:27
[10] Jan JC, Hsu WH, Liu SA, et al. Prognostic factors in patients with buccal squamous cell carcinoma: 10-year experience. J Oral Maxillofac Surg. 2011; 69 (2):396-404
[11] Kann BH, Buckstein M, Carpenter TJ, et al. Radiographic extracapsular extension and treatment outcomes in locally advanced oropharyngeal carcinoma. Head Neck. 2014; 36(12):1689-1694
[12] O'Sullivan B, Huang SH, Siu LL, et al. Deintensification candidate subgroups in human papillomavirus-related oropharyngeal cancer according to minimal risk of distant metastasis. J Clin Oncol. 2013; 31(5):543-550
[13] Aiken AH, Poliashenko S, Beitler JJ, et al. Accuracy of preoperative imaging in detecting nodal extracapsular spread in oral cavity squamous cell carcinoma. AJNR Am J Neuroradiol. 2015; 36(9):1776-1781
[14] Prabhu RS, Magliocca KR, Hanasoge S, et al. Accuracy of computed tomography for predicting pathologic nodal extracapsular extension in patients with head-and-neck cancer undergoing initial surgical resection. Int J Radiat Oncol Biol Phys. 2014; 88(1):122-129
[15] Chai RL, Rath TJ, Johnson JT, et al. Accuracy of computed tomography in the prediction of extracapsular spread of lymph node metastases in squamous cell carcinoma of the head and neck. JAMA Otolaryngol Head Neck Surg. 2013; 139(11):1187-1194
[16] McDermott M, Hughes M, Rath T, et al. Negative predictive value of surveillance PET/CT in head and neck squamous cell cancer. AJNR Am J Neuroradiol. 2013; 34(8):1632-1636
[17] Mehanna H, Wong W-L, McConkey CC, et al. PET-NECK Trial Management Group. PET-CT surveillance versus neck dissection in advanced head and neck cancer. N Engl J Med. 2016; 374(15):1444-1454
[18] Arellano K, Mosley JC, III. Case report of ipilimumab-induced diffuse, nonnecrotizing granulomatous lymphadenitis and granulomatous vasculitis. J Pharm Pract. 2017:897190017699762
[19] van den Brekel MW, van der Waal I, Meijer CJ, Freeman JL, Castelijns JA, Snow GB. The incidence of micrometastases in neck dissection specimens obtained from elective neck dissections. Laryngoscope. 1996; 106 (8):987-991