INTRODUCTION — Management of the neck in patients with metastatic involvement of cervical lymph nodes from a mucosal primary head and neck squamous cell carcinoma (HNSCC) has evolved with the increasing use of definitive radiotherapy (RT), either alone or in combination with chemotherapy. The neck is typically irradiated in continuity with the primary site if there is clinical/radiologic evidence of nodal disease or if the perceived risk of subclinical disease (N0) in the neck exceeds 15 to 20 percent. (See "Locally advanced squamous cell carcinoma of the head and neck: Approaches combining chemotherapy and radiation therapy".)
Traditionally, up to 40 to 50 percent of node positive patients with HNSCC managed with RT alone had persistent nodal disease on pathology, which provided the rationale for a planned neck dissection following definitive RT [1,2]. However, this approach meant that a substantial number of patients would undergo unnecessary surgery, with its associated morbidity.
The management of node positive HNSCC in patients who achieve a complete response at the primary site following RT is discussed here. Recommendations for treatment according to the primary site of disease are presented separately:
●(See "Treatment of locoregionally advanced (stage III and IV) head and neck cancer: The oral cavity".)
●(See "Overview of the treatment of locoregionally advanced head and neck cancer: The oropharynx".)
RATIONALE — When radiotherapy (RT) is followed by a consolidation neck dissection, there is an increase in complication rates and the severity of late toxicity, which can lead to an inferior quality of life in patients with node positive HNSCC [3-6]. Reducing the rate of severe late toxicity and its burden on survivors is increasingly important in this era of improved cure rates and prolonged survival following definitive RT. (See "Management of late complications of head and neck cancer and its treatment".)
However, many of these patients will not relapse if surgery is withheld. Thus, routine neck dissection imposes excess morbidity on these patients, and careful observation is the preferred approach for patients achieving a complete nodal response. If a complete response is achieved, the rate of isolated nodal failures remains low, regardless of initial nodal size, even though larger nodes are generally less likely to achieve a complete response [7,8].
Older criteria used to select patients for neck dissection included routine posttherapy neck dissection, regardless of neck response, for those with pretreatment disease >N1 (table 1A and table 1B and table 2 and table 3); for those who achieved less than a complete response (residuum >1 cm) in the neck; and for all those with N3 disease, regardless of neck response. The isolated nodal failure rate following a complete clinical and radiologic (<1 to 1.5 cm) response is less than 5 percent [7]. Advances in posttreatment imaging, improvements in RT techniques and the use of chemoradiotherapy, and changes in the epidemiology of HNSCC have led to major changes to the approach following RT or chemoradiotherapy. (See 'Restaging investigations' below and "General principles of radiation therapy for head and neck cancer".)
Epidemiologic changes in HNSCC, particularly throughout the Western world, now include an increasing number of patients with human papillomavirus (HPV) associated oropharyngeal squamous cell carcinoma and a decreasing number of cases with smoking-related tumors [9]. Patients with HPV-associated HNSCC tend to be younger, are former or never smokers, and have a significantly better prognosis (table 4). In addition, their mortality from second primary tumors or intercurrent disease is very low. (See "Epidemiology, staging, and clinical presentation of human papillomavirus associated head and neck cancer".)
Their likelihood of long-term survival following successful treatment makes reduction of late treatment-related toxicity an imperative. Coupled with improved locoregional control rates due to improvements in RT techniques, the likelihood of persistent pathologic neck disease following RT, even in the presence of residual nodal abnormalities on imaging, is lower than historical outcomes [10].
GENERAL APPROACH — Structural imaging (computed tomography [CT] and/or magnetic resonance imaging [MRI]) and functional imaging with F-18-fluorodexoyglucose (FDG) positron emission tomography (PET) have significantly improved the assessment of response status in the neck following radiotherapy (RT). (See 'Structural imaging' below and 'Functional imaging' below.)
The information from these studies is integrated with the clinical evaluation of the response, as well as the pretreatment assessment of the likelihood of recurrence, to formulate a plan for follow-up. The goal is to intervene surgically, if there is clear evidence of residual or recurrent disease, while distinguishing observed abnormalities from potential treatment-related changes.
The results of clinical assessment, structural and functional imaging, and consideration of clinical factors help to define indications for observation or surgery (algorithm 1). This approach is generally consistent with that proposed by the American Society of Clinical Oncology (ASCO) and the National Comprehensive Cancer Network (NCCN) [11-13].
Observation — Observation is preferred over surgery in the absence of evidence of residual disease or an increased risk of recurrence:
●Absence of evidence of residual disease by 12 weeks, as defined below:
•Complete clinical/structural imaging (<1 to 1.5 cm) response by 12 weeks.
•PET negative/structural imaging negative for nodal disease at 12 weeks.
●Ongoing regression or stable clinical/radiologic disease at 12 weeks:
•Ongoing clinical/structural imaging nodal regression beyond 12 weeks until <1 to 1.5 cm or stable residuum for at least six months (repeat imaging every two to three months).
•PET negative/structural imaging stable or regressing nodal disease at 12 weeks.
●A patient with equivocal PET findings and stable or regressing structural nodal abnormalities could have a repeat PET in four to six weeks; however, close clinical follow-up in four weeks and PET/CT eight weeks later are reasonable in situations in which the residual abnormality is thought to be nonviable disease. Subsequent management should be based on these findings.
Neck dissection — Indications for a neck dissection include the following:
●Clinical/structural imaging nodal progression at any time point following treatment.
●No nodal response by clinical and structural imaging >4 to 6 weeks posttherapy.
●PET positive/CT positive nodal disease 12 weeks posttherapy.
●In patients with gross lower neck disease (level IV and VA), administering a radiation dose (>60 Gy) may increase the risk of late brachial plexopathy. Therefore, limiting the dose of radiation to the lower neck to 60 to 66 Gy may be advisable. In these instances, a "planned" lower neck dissection, when combined with a lower dose of radiation, can be considered.
The threshold to perform a neck dissection is lower in patients where the likelihood of residual disease is high, such as suboptimally treated patients due to premature cessation of treatment, unplanned treatment interruptions, or the presence of high-risk primary disease, such as nonoropharyngeal HNSCC or human papillomavirus (HPV) negative oropharyngeal cancer.
Extent of neck dissection — The extent of the neck dissection should be tailored to the residual nodal disease and the neck level(s) involved. RT is effective in more than 95 percent of cases for sterilizing subclinical disease in the neck. Based on this principle, posttreatment neck dissection focuses on the areas suspected of harboring residual disease.
For example, if a patient is treated for a T2N1 HPV-associated tonsil cancer (table 1A-B) and is suspected of harboring residual disease in level II, based on imaging, a focused neck dissection may be performed. In this instance, neck levels IIA, IIB, and III would be adequate to remove the levels at highest risk for harboring residual viable tumor [14,15]. More extensive neck dissections do not improve regional control or survival but will increase morbidity.
Planned surgery versus PET/CT surveillance — The use of observation, rather than surgery, for appropriately selected patients is supported by the results of a large randomized trial demonstrating noninferiority in terms of overall survival and decreased rates of surgical complications [16]. In this trial, 564 patients with advanced HNSCC (N2 or N3 nodal disease) were randomly assigned to either planned neck dissection or surveillance including positron emission tomography (PET)/computed tomography (CT) at 12 weeks following initial definitive therapy. The primary outcome of the trial was noninferiority based on two-year overall survival; median follow-up was 36 months.
All patients were initially managed with RT (including concurrent platinum chemotherapy in 95 percent). The primary tumor was oropharyngeal in 84 percent of cases, and 74 percent were either current or former smokers. HPV status was assessed in 79 percent of patients (based on p16 positivity), and approximately 75 percent of those tested were p16 positive.
●Of the patients assigned to planned surgery, 78 percent actually underwent neck dissection. The remainder either declined surgery or were no longer surgical candidates following RT or chemoradiotherapy.
●Of the 270 patients assigned to surveillance, 96 percent underwent PET/CT at 12 weeks according to protocol. Complete imaging responses (functional and structural) were observed in 185 (69 percent); 15 had a complete response in the neck but not the primary (6 percent), and 66 did not have a complete response in the neck (24 percent). Of those with an incomplete response in the neck, 47 of 66 subsequently underwent a neck dissection; only those who were medically unfit or who had residual disease at the primary site did not undergo neck dissection.
●The two-year overall survival rates were 81.5 and 84.9 percent in the planned surgery and surveillance groups (hazard ratio 0.92 for surveillance, 95% CI 0.65-1.32). This met the predefined standard for noninferiority and excluded an unfavorable difference in two-year overall survival of more than 4 percent.
●Surgical complications were observed in 83 of those undergoing a planned neck dissection (38 percent). For those originally assigned to observation, surgical complications were observed in 22 cases (42 percent).
●There was no significant difference between planned surgery and surveillance in those with p16 positive tumors and those with p16 negative tumors. However, the overall prognosis for those with p16 positive tumors was significantly better than that for those with HPV negative tumors, consistent with other reports. (See "Epidemiology, staging, and clinical presentation of human papillomavirus associated head and neck cancer".)
RESTAGING INVESTIGATIONS — Determining whether a complete pathologic response has occurred can be difficult to assess in patients with node positive HNSCC due to fibrosis and residual nodal abnormality associated with the disease and its treatment. Key issues for restaging include the choice of investigation and the timing of its application.
Structural imaging — Structural imaging can be performed with computed tomography (CT), magnetic resonance imaging (MRI), or ultrasonography. CT is generally preferred, but MRI may be favored in cases where it is being used to reassess the primary site, such as nasopharyngeal or base of tongue tumors. Ultrasound is the cheapest and least invasive modality, and it can provide information on nodal size, architecture, and blood flow. Dense, fibrotic, and avascular lymph nodes on ultrasound imaging in general do not contain residual tumor.
Typically, the initial restaging scan should be performed around 12 weeks posttherapy, assuming that there is ongoing clinical response. Scans performed earlier tend to have higher rates of false positives. Delayed assessment may result in missing the optimal therapeutic window when a neck dissection is required, due to the onset of fibrosis and/or tumor progression. Up to 50 percent of patients will have a residual nodal abnormality based on CT criteria following radiotherapy (RT) in node positive HNSCC, but not all residual nodes will harbor viable disease [17,18].
In a study of 880 patients treated for node positive HNSCC with RT or chemoradiotherapy, 268 patients had a neck dissection for a residual structural nodal abnormality following RT; of these, 30 percent were found to have viable tumor, but only 4 percent had a subsequent isolated nodal failure [17]. This rate was similar to that in the 355 patients who achieved a complete nodal response based on structural imaging and did not undergo a lymph node dissection. These results imply that approximately 70 percent of patients underwent an unnecessary neck dissection.
Functional imaging — F-18-fluordeoxyglucose (FDG) positron emission tomography (PET) is the most commonly utilized functional imaging modality to assess the neck. FDG-PET is increasingly being used in combination with CT, providing synchronous functional and anatomic information.
A multicenter trial evaluated the negative predictive value (NPV) of combined PET/CT in clinically N0 necks, defined as being free of palpable lymph nodes with lymph node size on neck CT and/or neck MRI of <1 to 1.5 cm for jugular digastric nodes (IIa), spinal accessory nodes (IIb), or submental-submandibular nodes (Ia and Ib) or showing a lack of central lymph node necrosis in nodes of any size [19]. This trial included 212 patients with T2 to T4 squamous cell carcinoma of the oral cavity (64 percent), pharynx (17 percent), or glottis (19 percent) with at least one clinically N0 neck side who underwent neck dissection for pathologic confirmation. In this population, negative PET/CT scans were true negative in 87 percent and false negative in 13 percent. In contrast, positive PET/CT scans were true positive in 44 percent and false positive in 56 percent. The NPV varied depending on whether positivity was identified using a visual assessment (NPV 0.87) or a maximum standardized uptake value (SUV) cutoff (NPV 0.94 with a cutoff of 1.8).
A meta-analysis of earlier studies demonstrated the superiority of PET compared with CT in the pretreatment assessment of the neck [20]. A systematic review and meta-analysis of PET following RT found that the NPV of PET for the neck was >95, similar to the NPV of CT following a complete response [21,22].
While the NPV for CT is lower in the presence of residual nodal abnormalities >1 to 1.5 cm, the NPV of PET for residual nodal abnormalities based on either neck dissections or long-term observation was 95 to 97 percent [23,24]. The optimal timing for the restaging PET seemed to be between 8 and 12 weeks. However, multiple series have reported a relatively high frequency of false positives, predominantly due to ongoing inflammatory changes, with a positive predictive value (PPV) between 20 and 50 percent for PET scans earlier than 12 weeks [19,22-25].
Impact of changing biology of HNSCC — The sensitivities, specificities, and predictive values of any diagnostic imaging depend on the probability that there is residual disease. Where the likelihood of tumor response to RT is increased, such as in human papillomavirus (HPV) associated oropharyngeal squamous cell carcinoma, the NPV of the test may be increased. While the 12-week posttherapy PET/CT for HPV associated disease had a high NPV of 91 to 93 percent, the sensitivity and PPV remained low [10]. Further evaluation is needed [18,22].
Retrospective data suggest that for patients with HPV associated oropharyngeal squamous cell carcinoma, the decision to perform a posttherapy neck dissection can be deferred to 16 weeks without compromise in cure. This was illustrated in an analysis of 235 patients with HPV associated oropharyngeal cancer who achieved a complete response at the primary site following chemoradiotherapy and had no evidence of a distant relapse or nodal progression on clinical examination [26]. PET restaging at 12 weeks identified a complete metabolic response in 194 (82 percent). The 41 patients who did not achieve a complete response in the neck were observed, and 29 (71 percent) demonstrated a complete metabolic response on repeat PET at 16 weeks. Of the 12 with an incomplete response at 16 weeks, six proceeded with posttherapy neck dissection (four were pathologically positive), and six were followed clinically with no subsequent nodal failures.
In addition, the rate of posttherapy neck dissection has declined, likely due to the increased chemoradiosensitivity of HPV-associated oropharyngeal squamous cell carcinoma and improved restaging accuracy with FDG PET-CT. In a study of 418 patients with node-positive HPV-associated oropharyngeal disease, only 4.5 percent required a posttherapy neck dissection rate with a five-year nodal failure-free survival of 93.4 percent [27].
Emerging imaging techniques — While FDG has been the most commonly used radiopharmaceutical, there are a number of others in development, such as 3-deoxy-3-(18)F-fluorothymidine (FLT), but none that has entered routine practice.
Diffusion-weighted MRI has been studied in the post-RT setting to detect residual nodal disease at an earlier time point than PET or CT imaging, but further studies are required to confirm its utility [25,28].
Fine needle aspiration and cytology — While pretherapy image-guided fine needle aspiration (FNA) and cytology of a neck node for the detection of disease are relatively accurate, posttherapy cytology is less reliable due to the difficulty of interpreting the viability of tumor cells that have been irradiated.
In a study of patients undergoing neck dissection for residual/recurrent neck disease following curative RT, only 40 percent of patients who had positive cytology were found to have viable disease, while 19 percent of patients who had negative cytology turned out to have residual disease [29]. Thus, cytology from FNA cannot be solely relied upon to direct the posttherapy management of the neck.
MANAGEMENT OF A RESIDUAL NODE — The decision to observe the neck or perform surgery depends on response and likelihood that there is residual disease. While there is justification for observation following a complete nodal response on structural imaging, determining the management of patients with a residual nodal abnormality remains a dilemma.
F-18-fluordeoxyglucose (FDG) positron emission tomography (PET) is an important adjunct to structural imaging in the evaluation of the posttreatment neck. A negative assessment of a residual node using FDG-PET appears to be useful, but uncertainty remains when the node is positive by PET, and some of the data are conflicting:
●A prospective study of PET-directed management for patients following radiotherapy (RT) for node positive HNSCC provides evidence supporting the value of a negative PET [18]. In this study, 112 consecutive patients were managed with RT or chemoradiotherapy and underwent computed tomography (CT) and PET evaluation 12 weeks after therapy. Of the 50 patients with residual abnormalities on CT, 41 were negative by PET. At a median follow-up of 28 months, there were no isolated neck recurrences in these 41 cases. For patients who had equivocal PET nodal findings at 12 weeks, a repeat PET was performed four to six weeks later to improve the accuracy of the PET. For this cohort, if management had been based on a policy of neck dissection in the presence of residual CT nodal abnormalities, 44 percent of patients would have undergone surgery, compared with 7 percent using PET-directed management.
●PET appears to predict complete response earlier than CT as it can take several months for a residual node to regress to <1 cm. In a study by the Trans Tasman Radiation Oncology Group (TROG 98.02), patients with N2/3 disease who achieved a complete response at the primary site following RT were assessed radiologically at 12 weeks. For patients with residual CT nodal abnormalities at 12 weeks and a negative PET, observation was continued, and a repeat CT scan was performed at 26 weeks. There were no isolated nodal failures in patients who ultimately achieved a complete radiologic response in the neck [30].
●In contrast, in another study of 98 patients, the use of PET in addition to CT at eight weeks after treatment provided minimal additional information beyond that from the CT [22].
Data suggest that treatment response should continue to be made on the basis of the clinical context and qualitative parameters. In one study, investigators examined the role of standardized uptake values (SUV) to predict response to therapy. With maximum SUV cut-off values of 2.5 to 3, there are reports of high sensitivity and specificity for residual nodal disease. However, there are a number of factors (such as the presence of necrosis and other physiologic variations) that can make interpretation difficult [18]. Another study compared four qualitative interpretative criteria to assess posttreatment FDG PET-CT and found that all four had similar diagnostic criteria but that the Porceddu and Deauville scores offered the best trade-off of minimizing indeterminate outcomes while maintaining a high negative predictive value (NPV) [31].
LONG-TERM SURVEILLANCE — The goals of long-term surveillance following the decision to either observe or perform a neck dissection around 10 to 12 weeks posttherapy include the following:
●Early detection of locoregional recurrence, which may be amendable to curative salvage surgery
●Detection of second primary cancers
●Management of late effects
The majority of recurrences in the primary site and/or neck occur in the first three years following treatment and less frequently beyond that, up to five years. The general aspects of surveillance following treatment of HNSCC are discussed separately. (See "Posttreatment surveillance of squamous cell carcinoma of the head and neck".)
While regular surveillance of the neck with ultrasound or computed tomography (CT) may lead to early detection of a nodal recurrence or distant disease, there is a lack of evidence that this translates into survival, functional, or cost benefits. In the presence of a residual structural imaging nodal abnormality, three monthly ultrasound or CT imaging evaluations are performed until the residuum has resolved to <1 to 1.5 cm or the abnormality is stable for >6 months.
However, indiscriminate use of routine imaging may result in higher rates of false positive findings, leading to unnecessary investigations, anxiety, and increased health costs [32-34]. Hence, routine imaging surveillance is individualized following a discussion of the evidence with the patient.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Head and neck cancer".)
SUMMARY
●The optimal management of patients with head and neck squamous cell carcinoma (HNSCC) and clinical involvement of cervical lymph nodes following radiotherapy (RT) with or without chemotherapy is evolving. Key factors include improvements in treatment, advances in imaging, and epidemiologic changes in HNSCC. Routine consolidation neck dissection to remove nodes originally involved has been replaced by observation in patients who achieve a complete response in the primary tumor and cervical lymph nodes. (See 'Rationale' above.)
●F-18-fluordeoxyglucose (FDG) positron emission tomography (PET) has added functional imaging to the previously utilized structural techniques (computed tomography [CT]/magnetic resonance imaging [MRI]). A negative PET has a high negative predictive value, and a negative PET is an important indicator that disease is not present and is associated with a low incidence of isolated neck recurrence. (See 'Functional imaging' above and 'Management of a residual node' above.)
●The results of clinical evaluation, structural imaging (CT/MRI), and functional imaging (PET) are complementary (algorithm 1). This information can be used to identify populations that are at low risk for isolated neck recurrence and, hence, can be observed without further treatment, as well as those patients who should undergo neck dissection. (See 'General approach' above.)
●For patients in whom salvage surgery is indicated, the extent of the neck dissection should be tailored to the residual nodal disease and the neck level(s) involved. RT is highly effective for sterilizing subclinical disease in the neck, and posttreatment neck dissection only focuses on the areas suspected of harboring residual disease. (See 'Extent of neck dissection' above.)
