Extracapsular Spread: Biology, Imaging, Surgical Management, Pathological Diagnosis, and Therapeutic Implications

5 Extracapsular Spread: Biology, Imaging, Surgical Management, Pathological Diagnosis, and Therapeutic Implications

Kathryn M. Vorwald and J. Trad Wadsworth


This chapter is a comprehensive discussion of the most current knowledge of a familiar but ever-evolving topic in head and neck cancer. The recent addition of extracapsular spread (ECS) as a criterion used in the eighth edition of the American Joint Committee on Cancer (AJCC) TNM (t umor size, n ode involvement, and m etastasis status) staging system reflects the importance of ECS on prognosis and will undoubtedly produce an abundance of updated clinical data. This chapter encompasses everything from diagnosis to management and describes the current and future investigations that will continue to change the landscape in each of these areas.

Keywords: extracapsular spread, extracapsular extension, extranodal spread, extranodal extension, perinodal spread, perinodal extension, transnodal spread, transnodal extension, cervical nodal metastasis, high-risk head and neck cancer

5.1 Background of Extracapsular Spread

Extracapsular spread (ECS) of metastatic head and neck cancer (HNC) continues to capture the interest of professionals today just as it did for Rupert A. Willis, MD, back in 1928 when he encountered the impressive dissemination of “epidermoid carcinoma” of the head and neck during his postmortem examinations.1 ECS refers to extension of malignant tumor from within a lymph node through the fibrous capsule and into the surrounding connective tissue. There are various terms that have been used to refer to ECS including extracapsular extension, extranodal spread/extension, perinodal spread/extension, transcapsular spread/extension, and capsular rupture. It also includes those soft-tissue deposits of carcinoma found in the tissues of the neck that lack evidence of lymphoid tissue but are not a direct extension of the primary tumor as these may represent extralymphatic deposits of carcinoma or simply lymph nodes that have been completely replaced by tumor.2

ECS eventually found its way into the forefront of HNC in 1971 when Bennett et al sought to identify additional prognostic factors. They found that the presence of ECS resulted in a decline in survival rates beyond that of regional metastasis alone.2 Early investigators thus urged for its consideration as a prognostic criterion to place patients into a high-risk category.3 As such, they recommended considering adjuvant therapies for those patients with ECS in hopes of achieving better local and regional control to improve cure rates.

An abundance of research has since been conducted in relation to ECS, confirming its profoundly negative effect on outcomes of HNC with a 5-year disease-specific survival (DSS) of a mere 25.0% in patients with ECS as compared to those without at 57.8% (p < 0.001).4 This remained statistically significant even in multivariate models that accounted for the effect of other prognostic factors. Furthermore, there was found to be over twice the risk of death from HNC in those with ECS present, based on a hazard ratio of 2.44. As a result of research such as this, ECS has now earned its place in the formal staging of HNCs. The eighth edition of the American Joint Committee on Cancer (AJCC) Staging Manual5 now includes ECS in its nodal classification, which preliminary evidence demonstrates will more accurately reflect the true prognosis of HNCs.6,7,8 This is due to the upstaging of many cancers as a direct result of the presence of ECS-producing N classifications of N2a or N3b, categorizing them as stage IV cancers.5 This will also likely have the effect of increasing the use of adjuvant therapies, accordingly.

Inclusion of the ECS criterion applies to the majority of HNCs but does not apply to human papillomavirus (HPV) related oropharyngeal cancers or mucosal melanomas. The HPV-positive oropharyngeal cancers were recently separated due to insight into their biologic difference and subsequent improvement in prognosis, regardless of ECS status.6 From this point forward, the chapter will refer to ECS as it pertains to HPV-negative squamous cell carcinoma, unless otherwise specified.

Even with this noteworthy progress, ECS is still in its infancy of discovery, and much of the historical data may be somewhat compromised. This can be partially attributed to the nonexistence of defined clinical and pathological diagnostic criteria, an absence of specific pathologic reporting, and a lack of HPV designation. The reported incidences of ECS can also vary due to compounding effects of other factors such as procedure performed, interpreting pathologist, primary site, lymph node size, disease stage, prior therapies, and tumor biology. Regardless of the variation, there is undoubtedly a high overall incidence of ECS. In a recent study of nearly 300 patients with oral cancer, there was found to be an overall incidence of ECS of 54.1% in patients with regional metastases,8 which was similar to a large meta-analysis of 1,188 patients with an overall incidence of 50.5%.7 Even in clinically negative (cN0) necks, ECS has been microscopically identified in 22.2% of patients.9 With such a high prevalence and new role in nodal staging, there is certain to be continued growth surrounding this topic.

5.2 Biology of Extracapsular Spread

Despite a wealth of clinical data on the impact of ECS on patients with HNC, there seems to be a relative paucity of knowledge on the actual pathophysiologic mechanism behind it. The complex nature of cancer biology, in particular how cancer cells are able to migrate outside of the lymph node capsule, still riddles us today. It was initially thought to be simply due to the outgrowth of the enlarging metastatic tumor beyond the lymph node boundaries, and studies have, in fact, demonstrated an association between node size and ECS.4,8 However, the fact that even subcentimeter nodes with tiny deposits of tumor also exhibit ECS10 challenges this idea and calls for alternative theories.

Recently, Curry et al built off the foundation of the well-known monocarboxylate transporter 4 (MCT4) and its association with cellular motility and invasiveness in various cancer types including head and neck tumors.11 They demonstrated its presence in the cancer-associated fibroblasts found in the extracellular matrix surrounding primary HNCs as well as those specific sites where tumor cells have broken through the lymph node capsule. Therefore, they believe that the MCT4 may have a significant role in HNC invasion, including that which occurs in ECS. However, further investigation is needed to determine causality. There has also been extensive work on the epithelial-mesenchymal transition of cells, which has been shown to allow for invasion and metastatic dissemination of cancer cells through increased motility and invasiveness by degradation of the extracellular matrix.12 It can be postulated that these same mechanisms at work in the primary tumor cells are also present in those which have metastasized to lymph nodes and potentially permit ECS.13

These ideas and the rampant search for potential biomarkers to assist in predicting and identifying ECS require continued study but may, in the future, provide valuable information in determining the actual mechanism by which ECS portends a worse prognosis and potentially give insight into developing therapies. It may also aid in the diagnosis and study of other ECS characteristics such as distance of spread beyond the capsule, level of the neck where ECS is identified, total number of nodes with ECS, or the ratio of metastatic lymph nodes to ECS. As the pursuit of demystifying the science behind ECS continues, the clinical management of ECS must, however, carry on and adapt to rising evidence.

5.3 Diagnosing Extracapsular Spread with Imaging

The first step in managing ECS is identifying its presence, so studies have investigated clinicians’ abilities to utilize various imaging modalities to detect ECS. There have been several characteristics reported for its radiological diagnosis, and they are similar in both contrast-enhanced computed tomography (CT) and magnetic resonance imaging (MRI). Fujita et al have described the “established criteria” for radiographic evidence of macroscopic ECS to include thick-walled enhancing nodal margins, loss of outer nodal margin definition, capsular contour irregularity, infiltration of adjacent fat planes around portions of the node, invasion of adjacent structures, and multiple nodes abutting one another with loss of an intervening fat plane.14

The use of contrast-enhanced CT has shown a lack of specificity and sensitivity in correctly identifying ECS through simple interpretation of the images alone. Therefore, attempts have been made to improve the accuracy of diagnosis by combining associated findings such as central nodal necrosis (image Fig. 5.1), which raised the sensitivity and specificity to 95 and 85%, respectively,15 and has been found to be an independent radiologic predictor of ECS.16,17 Specificity was improved to 94% with the identification of three or more nodes with radiographically suspicious characteristics including size greater than 2 cm in a single dimension, heterogeneous appearance, cystic change, and/or irregular borders.18

MRI is often touted as having better soft-tissue resolution, in general; however, this advantage has not been apparent in the diagnosis of ECS. In fact, the use of MRI to predict ECS was previously found to be no better than clinical palpation alone, both of which were inadequate as compared to the actual pathologic rate of ECS.19 A recent systematic review and meta-analysis found that, although MRI may have a significantly higher sensitivity than CT, there was no difference in specificity and both had similar diagnostic efficacies.20 One group identified a unique approach utilizing physiologic differences in lymph node blood flow and MRI signal, developing a pixel-based time-signal intensity curve profile within nodes 10 mm or larger and found a specificity of 100% and sensitivity of 96%, but this has not been studied further for application in the everyday clinical setting.21

Ultrasound examination is able to utilize additional characteristics such as marked internal echogenicity and the absence or narrowing of hilar echoes which, in combination with the known characteristics of ECS, further suggests its presence.22 Despite this, no significant differences in the sensitivity and specificity have been found between ultrasound and CT.23 In a comparison of ultrasound and MRI, they were found to have comparable accuracies, but ultrasound had a higher specificity than MRI.24

Given the shortcomings of standard imaging modalities, there was hope that the addition of metabolic information from fluorodeoxyglucose positron emission tomography (FDG PET) would enhance the ability to diagnose ECS. In fact, several retrospective studies from the same institution utilized FDG PET/CT images to deduct threshold SUVmax values ranging from 2.25 to 3.85 to identify ECS from various HNC primary subsites.25 Unfortunately, sensitivities and specificities using these values ranged from a disappointing 80 to 85.7% and 74 to 88%, respectively. Others have actually demonstrated that ECS of HNC is more accurately detected by simply using contrast-enhanced CT alone rather than FDG PET/CT.26

Despite ever-improving technology and resolution, the AJCC Head and Neck Task Force determined that today’s imaging modalities do not have the level of sensitivity and specificity to rely upon their ability to identify early or minor ECS and, therefore, cannot be the sole basis for diagnosing ECS when applying it to the AJCC clinical staging.5 They do, however, allow for radiologic imaging as supportive evidence to physical examination in order to diagnose it, but the physical signs must then be unambiguous, such as with skin invasion, infiltration or dense tethering of musculature or adjacent structures, or dysfunction of a cranial nerve/brachial plexus/sympathetic trunk/phrenic nerve (image Fig. 5.2). This decision is further supported by the meta-analysis performed by Su et al20 where they assessed the pooled data of 15 different studies and found a mean sensitivity/specificity for CT of 0.77/0.85, for MRI of 0.85/0.84, and for PET/CT of 0.86/0.86. Overall, this confirms that imaging modalities do not yet have the ability to provide routine, consistently accurate clinical diagnoses. This, in combination with the lack of standard diagnostic criteria, calls for the development of improved methods of detection since ECS has been shown to have a major role in prognosis and treatment decision-making. At the current time, it appears that the contrast-enhanced high-resolution thin-slice CT scan is the most valuable imaging modality for use as a routine study to predict ECS in a nonsurgical manner. Despite its shortcomings, it may provide additional information otherwise missed by physical examination alone.

5.4 Surgical Management of the Neck

Since radiographic studies alone are not currently capable of reliably diagnosing ECS, surgical excision remains the only method to allow for histopathologic examination, which is the gold standard for diagnosis. The surgical management is the same for the clinically positive neck, namely, a modified radical neck dissection. However, the question that remains is how to address the cN0 neck, particularly in the new staging era of ECS inclusion. As with any surgical procedure, the decision relies upon assessment of the risks versus benefits. The general risk of performing a selective neck dissection (SND) is present and described elsewhere in this text but is of low morbidity in the hands of an experienced surgeon. The functional and esthetic results after an SND are also typically acceptable. This risk may be deemed worthwhile, if occult disease is identified; however, the ultimate risk would be in performing a neck dissection found to be pathologically negative (pN0), having then performed a truly nontherapeutic surgical intervention.

On the other hand, the benefit of having complete pathologic classification may be considered worthwhile, regardless of the findings. This would allow for proper staging upfront and avoid the potential of understaging, given the fact that minor ECS is equivalent to major ECS in the new system.5 Even a pN0 classification is beneficial, particularly if it allows for de-escalation of adjuvant therapies. The additional benefit of identifying occult metastases, and especially occult ECS, is the ability to intervene early on in the process as compared to late where regional relapses in the undissected neck have been found to present at a more advanced stage and with higher rates of ECS.27

There is evidence to support the need to perform elective neck dissections based on the rate of occult ECS. In a 2002 study by Coatesworth and MacLennan,9 22.2% of all cN0 necks were found to have microscopic ECS and/or soft-tissue deposits present. Nearly 75% of the pathologically positive (pN +) necks had microscopic ECS and/or soft-tissue deposits present. In another study by the same group, 25% of cN0 necks examined and 44% of cN1 necks had ECS and/or soft-tissue deposits.28 This is evidence that microscopic ECS and soft-tissue deposits are highly prevalent in patients with cN0 necks and that ECS can occur even at an early stage in metastasis. To highlight the influence of ECS, they also found no significant difference in the outcomes of patients with a pN0 neck and those with a pN + neck but with completely encapsulated nodes (i.e., without ECS).29

More recently, D’Cruz et al27 found that 74% of the cT1/T2 oral cavity node-negative patients in the nonelective neck dissection group eventually developed recurrent disease in the neck, and these patients had a significantly higher incidence of ECS (p < 0.001) with a more advanced nodal stage (p = 0.005) as compared to those treated with an elective neck dissection. Furthermore, the patients who were found to have pN + neck disease in the upfront elective neck dissection had significantly better overall survival (OS) rates than those who had been carefully observed and later presented with nodal relapse.30

Similarly, in a retrospective chart review by Dik et al, management of the cN0 neck in patients with surgically resected early-stage (cT1/T2) oral squamous cell carcinoma were handled in either a “watchful waiting” approach or underwent an intentional upfront ipsilateral SND.31 They found that a significantly larger number of positive lymph nodes with ECS were found in the watchful waiting group after they recurred than in those who were found to have nodal disease following upfront SND. This also produced a significantly worse DSS as compared to those who underwent upfront SND.

Additionally, a recent study found 66.7% of patients with salvage neck dissections had nodes with ECS present, and the 5-year salvage-specific survival for those patients was 32.0%, compared to 77.2% in those without ECS (p = 0.0001).32 They noted that these patients with recurrent regional disease and ECS present had a 4.1 times higher risk of death as a consequence of the tumor as compared to those who did not have ECS present. The variable found to be most related to the presence of ECS in salvage neck dissections was the class of regional neck recurrence, with 19.2% of rpN1 necks with ECS, 75.9% of rpN2 necks, and 100% of rpN3 necks (classified according to the AJCC staging manual, seventh edition).

The evidence from these studies helps one to appreciate that an upfront neck dissection can have major implications on staging, prognosis, and adjuvant therapeutic options, especially if ECS is identified. On the same note, the risk of missing occult metastases and/or ECS seems to be worse than the risk of performing a neck dissection on a pN0 neck. Therefore, elective neck dissections should be strongly considered. Further study is needed regarding the prognostic effects of microscopic ECS and recurrent disease with ECS, now more than ever, for clarification on the surgical management of the cN0 neck.

In all cases of neck dissection, whether elective or therapeutic, and particularly when ECS is suspected, surgeons should be mindful of the fact that ECS may be present, and direct contact with tumor cells risks implantation/seeding. Respecting the fascial planes when possible also helps avoid direct contact with potential tumor cells present in the perinodal adipose tissue. Likewise, “node picking” or “node sampling” may miss microscopic metastases and soft-tissue deposits. Clinical suspicion for ECS pre- or intraoperatively should prompt the removal of any adjacent tissues that appear to potentially be involved such as the overlying skin or adjacent structures in the neck. Careful surgical technique with ECS in mind should help avoid the potential for leaving behind microscopic disease or inadvertently seeding tumor cells.

5.5 Pathological Diagnosis of Extracapsular Spread

Currently, histopathologic evaluation is the only method of definitively identifying ECS. Despite being the gold standard, there is a level of variability and subjectivity in its diagnosis, which has been demonstrated in several studies, even with experienced head and neck pathologists.33 A low level of interob-server, and even intraobserver, agreement in the pathologic assessment of ECS has been contributed to the difficulties in evaluating lymph nodes that have been fully replaced by tumor and are surrounded by a desmoplastic reaction, mimicking a capsule, causing some pathologists to score lymph nodes negative for ECS.34 Another debate occurs when tumor is present at the lymph node hilum, where there is naturally no true capsule present, requiring subjective determination as to whether this should be considered a disruption in the capsule due to ECS or simply a deposit in the hilum. Juxtacapsular extension, where tumor tissue grows into the capsule but not outside of it, can also cause some disagreement among pathologists. This highlights the reality of the challenges in histopathologic diagnosis of ECS even though at first glance it seems straightforward. This may be partially due to the lack of a set of standardized diagnostic criteria; however, several common characteristics are reportedly used for its identification including perinodal fat involvement (image Fig. 5.3); skeletal muscle, nerve, and thick-walled vessel involvement; tumor beyond the nodal capsule; and desmoplastic stromal reaction outside the node.33

Once defined systems of diagnosis, classification, and reporting for ECS are in place, it will better enable associated information to be identified and collected, such as extent of disease beyond the lymph node capsule, which is an important topic of study. It is known that ECS is a poor prognostic factor, but, currently, both microscopic and macroscopic ECS have been lumped together. Some evidence shows there is a correlation between the level of extension of ECS and prognosis so microscopic ECS may play a role, but the exact threshold distance that is truly significant remains undefined.

In a study on 266 patients with oral cancer, Greenberg et al found no difference in survival when comparing the extent of ECS of ≤ 2 and > 2 mm from the capsule of the lymph node,35 indicating there may be no difference in survival between microscopic and macroscopic ECS, allowing them to be grouped together. Jose et al likewise found that there was no significant difference in overall or recurrence-free survival between microscopic and macroscopic ECS.29 A recent study on a cohort of pathologically node-positive oral cancer patients better defined the threshold when they found that the adverse prognosis of ECS was only clinically relevant when it extended more than 1.7 mm beyond the nodal capsule.4

Feb 14, 2020 | Posted by in OTOLARYNGOLOGY | Comments Off on Extracapsular Spread: Biology, Imaging, Surgical Management, Pathological Diagnosis, and Therapeutic Implications

Full access? Get Clinical Tree

Get Clinical Tree app for offline access