Head and neck squamous cell carcinoma of unknown primary

INTRODUCTION — Head and neck squamous cell carcinoma (SCC) of unknown primary is defined as the presence of SCC in one or more lymph nodes within the head and neck region that are not solely in the supraclavicular region, without an identifiable primary tumor. While many head and neck cancer patients initially present with a neck mass, the mucosally based primary site of origin is usually identified over the course of the subsequent workup.

Thus, SCC of a truly unknown primary origin is a relatively rare entity in the head and neck region. Furthermore, increasing sensitivity of positron emission tomography (PET)/computed tomography (CT) and transoral lingual tonsillectomy have enabled detection of microscopic, hard-to-identify oropharynx cancers, which are often HPV-related, which has further lowered the rate of unknown primary SCC. Retrospective studies suggest that unknown primary SCC represents between 1 and 3 percent of new cases of head and neck SCC [1]. It is commonly thought that many SCC of unknown primary represents clinically occult oropharyngeal cancers. A contemporary single institution series of 75 patients with SCC of unknown primary found that 69 (90.5 percent) were positive for human papilloma virus (HPV) by either p16 positivity or in situ hybridization [2], although some older series have reported a lower incidence [3].

The special case of isolated supraclavicular lymphadenopathy, including those of squamous cell origin, almost always originates from cancer in the skin or a primary site beneath the clavicles (ie, lung, breast, colon, prostate, esophagus, cervix, ureteral, and lymphoma primaries) [4]. These cancers are associated with a different natural history and prognosis than cancers originating from the head and neck.

The management of head and neck SCC of unknown primary is directed at cure for most patients; cervical lymphadenopathy represents disease that is locoregionally advanced but is not distantly metastatic in most patients. This is in sharp distinction to cancers of unknown primary originating from below the clavicles, for which supraclavicular adenopathy may represent incurable distant metastasis.

The diagnosis and management of patients with SCC of unknown primary presumed to originate from the head and neck are presented here. The initial evaluation of a neck mass and the general approach to neoplasms of unknown primary origin are discussed separately.

●(See "Evaluation of a neck mass in adults".)

●(See "Overview of the classification and management of cancers of unknown primary site".)

●(See "Squamous cell carcinoma of unknown primary site".)

DIAGNOSTIC EVALUATION — The diagnosis of head and neck squamous cell carcinoma (SCC) of unknown primary (ie, SCC in a lymph node or in multiple lymph nodes) within the head and neck region but not solely in the supraclavicular region is a diagnosis of exclusion.

Patient management for head and neck SCC of unknown primary is based on a diagnosis that is rendered after a thorough evaluation is performed in an effort to identify a primary tumor. This evaluation should, at a minimum, include a comprehensive history and physical examination, high-quality contrast-enhanced diagnostic imaging, and panendoscopy under anesthesia.

History and examination — The first step in the evaluation of a patient with a malignant neck mass and suspected head and neck SCC is a comprehensive history and physical examination. The examination should focus on identifying a potential primary source for malignant spread to the neck.

The history should include notation of any previous malignancies, including any possibility of undiagnosed skin cancers. Many patients, especially older persons with a long dermatologic history, may forget to describe procedures addressing small or superficial skin lesions excised, frozen, or curetted within the past few years. Persons with immune compromise should be considered to be at elevated risk for a primary skin cancer. Further history should elicit any symptoms related to the head and neck that could indicate a mucosal site of origin, such as a globus sensation, epistaxis, pain, otalgia, odynophagia, dysphagia, hemoptysis, or hoarseness. (See "Overview of the diagnosis and staging of head and neck cancer", section on 'Clinical presentation'.)

The physical examination should encompass a thorough examination of the scalp and non-hair-bearing skin for evidence of cutaneous malignancy as well as for scars that may represent prior dermatologic procedures performed on a cutaneous neoplasm. Specialist dermatologic expertise can be consulted for examination and investigation of suspicious skin findings. Mucosal primaries are unlikely to travel to the parotid gland, external jugular chain, or perifacial lymph nodes, and a cutaneous malignancy should be suspected in these cases. P16 positivity in a lymph node does not rule out a cutaneous primary as it is known to be positive in up to 60 percent of cutaneous SCC and 50 percent of basal cell carcinomas [5,6]. (See "Cutaneous squamous cell carcinoma (cSCC): Clinical features and diagnosis" and "Treatment and prognosis of low-risk cutaneous squamous cell carcinoma (cSCC)".)

If a diagnosis of carcinoma has been obtained by fine needle aspiration (FNA), the specimen can be tested for viral etiology to help direct the search for the primary tumor. Most, although not all, nasopharyngeal and oropharyngeal SCC can be linked to the presence of Epstein Barr virus (EBV) or human papillomavirus (HPV), respectively; both can be detected on FNA specimens. (See 'Investigational studies' below.)

Generally, an excisional biopsy of the neck mass should be avoided. Excisional biopsy should be used as a diagnostic tool of last resort following complete radiographic imaging, less invasive pathologic investigations, and completion of staging.

Flexible fiberoptic laryngoscopy is essential in order to evaluate the sinonasal cavity, nasopharynx, pharyngeal walls, base of tongue, larynx, and hypopharynx. Oropharyngeal tumors can be difficult to identify on routine visual inspection since small tumors are easily concealed within normal-appearing lymphoid tissue in the tonsillar crypts and base of tongue.

Thus, it is important to palpate the tonsils and base of tongue in order to detect areas of irregular firmness or easy bleeding. However, most patients cannot tolerate a thorough examination of this type in an office setting.

Radiographic imaging — Computed tomography (CT) and/or magnetic resonance imaging (MRI) with contrast are usually the first line of imaging for patients with head and neck SCC of unknown primary. If cross-sectional CT and/or MRI do not readily identify a primary tumor, positron emission tomography (PET) with fluorodeoxyglucose (FDG) or integrated PET/CT should be performed prior to panendoscopy or biopsies.

The information obtained from such cross-sectional imaging may direct subsequent biopsies. However, if a biopsy, or deep examination or manipulation of the tissues is performed prior to PET/CT, the scan will show FDG uptake in those areas due to inflammation from the procedures, and the specificity of the imaging is decreased. Some practitioners will use PET/CT as the initial imaging step instead of cross-sectional CT or MRI; however, it is important to note that structural and anatomic information can be lost if the CT component of the PET/CT is of low resolution and/or performed without contrast.

A comprehensive review, as well as several subsequent studies, confirmed the efficacy of PET in the detection of primary tumors in patients with cervical metastases from an unknown primary [7-11]. PET detected approximately 25 percent of tumors that were not detected after a workup that did not include PET. PET also was able to detect additional sites of metastases in 27 percent of patients. Integrated PET/CT imaging appears to be significantly more sensitive than PET alone in identifying the primary site [12].

Staging surgical endoscopy — "Panendoscopy" (sometimes called "triple endoscopy") denotes a set of endoscopic examinations performed under anesthesia. A direct laryngoscopy should be performed under anesthesia to evaluate the oropharynx, larynx, and hypopharynx. A nasopharyngoscopy is also performed concurrently, and upper esophagoscopy should usually be performed as well, particularly if there is suspicion of hypopharyngeal involvement.

While the patient is sedated, the oral cavity should also be examined and palpated thoroughly. Deep palpation of the tonsils and base of tongue can be performed with higher yield while the patient is under anesthesia and the tissues are relaxed.

Directed biopsies should be performed for all areas of clinical suspicion, ie, those associated with mucosal irregularities, those that bleed easily, and any areas of potential abnormality that were detected on imaging.

When the primary site is still not obvious from the endoscopic examination under anesthesia, most experts recommend that unilateral or bilateral tonsillectomy should be performed on patients with adequate lymphoid tonsillar tissue in the region of the palatine tonsils as part of the initial workup. In two studies that included tonsillectomy in the initial evaluation, 26 to 35 percent of the cases of unknown primary were found to have a cancer in the ipsilateral tonsil [13,14].

●A retrospective comparison study demonstrated that the overall yield for identifying an occult primary tonsillar cancer was increased from 3 percent in 95 patients who had deep tonsillar biopsies to 30 percent in 27 patients who had diagnostic tonsillectomy [15].

●When bilateral tonsillectomy was performed in 41 patients presenting with head and neck SCC, a tonsil primary was found in both tonsils in two patients (one with contralateral carcinoma in situ) and in the contralateral tonsil in two patients [16].

These results are in keeping with older clinical data suggesting that approximately 10 percent of tonsillar cancers will have bilateral cervical node metastases [17]. However, these data were obtained in the era before contemporary high-resolution cross-sectional imaging and PET/CT.

The argument for bilateral tonsillectomy is that the incidence of increased complications related to bilateral tonsillectomy is minimal, with a low rate of complications of any kind from tonsillectomy, and there is a chance of missing a contralateral tonsillar primary if a bilateral removal is not done. Furthermore, there is a theoretical advantage that subsequent examinations for cancer surveillance may be facilitated by a greater symmetry on physical examination after bilateral tonsillectomy. However, because of a low observed rate of clinically apparent failure in the tonsillar tissues contralateral to an involved neck node, there are major centers where it is thought that the increased morbidity related to the contralateral tonsillectomy is not warranted.

Lastly, just as palatine tonsillectomy has shown high yield for small primaries not identified on physical examination and diagnostic imaging, newer research has shown that lingual tonsillectomy is similarly revealing [18]. Since the majority of SCC of unknown primary are HPV/p16 related, lingual tonsillectomy (removal of the lymphoid tissue on the surface of the base of the tongue) is attractive in that it provides another means for further investigation of the oropharyngeal area in addition to palatine tonsillectomy.

Contemporary retrospective studies indicate that using transoral robotic surgery (TORS) to perform a lingual tonsillectomy during workup for an unknown primary improves the ability to identify a primary [19-21]. In one study, 10 patients underwent lingual tonsillectomy after a thorough workup (PET/CT, panendoscopy, bilateral palatine tonsillectomy) failed to reveal a primary tumor. In 9 out of 10 of these patients, a primary was identified in the lingual tonsils [19]. In a more recent retrospective case series, TORS found the primary site in 72 percent of cases not identified on imaging [22]. In this series, 59 percent of identified primaries were located in the base of the tongue, and 39 percent were located in the tonsil. TORS also allows for surgical resection of an identified oropharyngeal primary cancer.

However, the value of TORS in the identification of disease that remains occult following diagnostic imaging, examination under anesthesia, and traditional tonsillectomy has not yet been rigorously established. Historically, suppression of primary emergence (ie, treatment of the occult primary) has been achieved using wide-field, low-dose radiation. Whether more precise identification and resection using TORS results in improved cost effectiveness and functional outcomes requires further study.

Investigational studies — Even after these extensive evaluations, the primary tumor is occasionally not found. Newer diagnostic procedures show potential in aiding in the identification of the primary tumor.

Detection of HPV or EBV in the FNA biopsy from a pathologic lymph node is very useful and could provide prognostic information:

●The majority of primary tumors that are identified among patients with head and neck SCC of unknown primary are located within the oropharynx (43 percent tonsil and 39 percent base of tongue) [11]. Furthermore, more than 70 percent of oropharyngeal SCC are HPV positive [23], and only a more limited percentage of tumors from non-oropharynx head and neck sites are HPV positive. (See "Epidemiology, staging, and clinical presentation of human papillomavirus associated head and neck cancer", section on 'Clinical presentation'.)

The gold standard for assessing HPV infection is in situ hybridization or polymerase chain reaction (PCR) to detect HPV DNA. Immunohistochemical analysis of p16 is a valuable and readily available, albeit imperfect, surrogate biomarker to identify tumors associated with HPV infection [24]. A positive finding of p16 is less specific than identification of HPV DNA; however, the concordance between p16 and HPV is better than 90 percent in the oropharynx. The finding of p16 or HPV in a lymph node biopsy may help to narrow down the site of origin and may assist in focusing the examination under anesthesia. (See "Epidemiology, staging, and clinical presentation of human papillomavirus associated head and neck cancer", section on 'Confirming HPV 16 positivity'.)

In the setting of a nondiagnostic examination under anesthesia, p16/HPV analysis of the lymph node is considered by some to be sufficient to make a treatment decision, specifically to avoid irradiation of the hypopharynx and larynx (and even the nasopharynx, if the entity of HPV associated nasopharyngeal SCC is not in question). Instead, RT can focus on the oropharynx, thereby avoiding the morbidity of irradiation of the entire head and neck mucosal axis. The overall clinical scenario should be considered. As an example, this is a reasonable strategy for a young non-smoker with no symptoms, whereas a smoker with multiple head and neck symptoms would be quite likely to harbor an occult head and neck primary in other sites.

●Similarly, detection of EBV genome sequences by PCR is a sensitive marker for the presence of nasopharyngeal carcinoma (NPC), and a fine needle aspirate found to contain EBV DNA strongly suggests the nasopharynx as the source. (See "Epidemiology, etiology, and diagnosis of nasopharyngeal carcinoma".)

However, the relatively high frequency of FNA aspirates that do not show evidence for HPV or EBV means that a nondiagnostic or negative result does not exclude an oropharyngeal or nasopharyngeal primary tumor [25].

Analysis of serum or plasma EBV DNA may be useful in identifying a primary tumor in Eastern Asian and Middle Eastern populations, which are demographically at higher risk for NPC.

New technologies under investigation may enhance the ability to diagnose malignant lesions in the upper aerodigestive tract. As an example, time-resolved laser-induced fluorescence spectroscopy is a promising noninvasive diagnostic technique, particularly in discriminating between normal and malignant tissue [26].

Staining of tissues removed at biopsy for HPV or EBV may help direct diagnostic evaluation. However, false negative results occur, and the results of such tests should be evaluated carefully when making the decision to exclude mucosal sites from irradiation. As examples, in rare cases EBV positive tumors can arise outside of the nasopharynx, and HPV positive tumors can arise outside of the tonsil and tongue base [27]. Tissues may contain evidence of these infections without actual manifestation of a virally related cancer.

STAGING SYSTEM — The eighth edition (2017) of the tumor, node, metastasis (TNM) system developed jointly by the American Joint Committee on Cancer (AJCC) and the Union for International Cancer Control (UICC) is used to stage patients with cervical lymph nodes and an unknown primary tumor of the head and neck. This classification system is presented in the table (table 1) [28]. The key change in this edition is the new staging system for p16 or human papillomavirus (HPV) related oropharyngeal cancer (table 2 and table 3). It reflects this biologically defined cohort's improved prognosis in response to standard therapies. Because the eighth edition staging system is newer, all treatment recommendations are based on research studies using the previous staging system. For this reason, the following sections on treatment and the summary of recommendations use staging following the rules of the seventh edition of the AJCC staging system.

TREATMENT

Overview — Achieving locoregional disease control in patients with head and neck squamous cell carcinoma (SCC) of unknown primary has two components: controlling the gross disease in the neck and ensuring that a primary cancer does not develop in potential sites of origin (primary emergence). Confounding the interpretation of success in the control of mucosal primary emergence is the high rate of second primary development in head and neck cancer.

Emergence of a second primary tumor, generally thought to occur in 10 percent of patients with a prior history of head and neck cancer, appears to be much less prevalent in patients who develop human papillomavirus (HPV)-related oropharyngeal cancer. However, the development of a second malignancy in any patient is a serious event since treatment options may be more limited. (See "Second primary malignancies in patients with head and neck cancers".)

The optimal treatment of patients with head and neck SCC of unknown primary is uncertain [29]. The most important factor influencing prognosis appears to be the clinical stage at diagnosis. In a systematic review that included 18 studies and 1726 patients, the five-year survival rate decreased progressively from 61 percent for those with N1 disease to 26 percent for those with N3 disease [30].

The two broad approaches to treatment include primary surgery (usually followed by radiation therapy [RT] or, in some cases, chemoradiotherapy) and upfront definitive RT, either alone or in combination with chemotherapy. There are no prospective studies comparing these approaches, and data on their relative effectiveness in achieving locoregional control and disease-free survival are limited to observational studies. Treatment must consider both the disease in the neck as well as the occult primary in the head and neck mucosa.

The following treatment recommendations are based on the seventh edition American Joint Committee on Cancer (AJCC) staging system. The purpose of the AJCC eighth edition stage group is to provide prognostic value and not necessarily to guide treatment decision-making. (See 'Staging system' above and "Treatment of human papillomavirus associated oropharyngeal cancer", section on 'Staging'.)

●For patients with N1 disease without extracapsular extension (table 1), we suggest single-modality therapy (surgery or radiation). Many authors feel that low-volume N2a disease without any evidence of extracapsular extension may be handled in a similar fashion, although the data are more limited [31,32]. (See 'Primary surgery' below and 'Definitive radiation therapy' below.)

For patients with N1 disease without extracapsular extension, surgery can be performed, and postoperative RT may be withheld following a formal neck dissection, although these patients should be offered a discussion of the risks and benefits of RT to mucosal regions (toxicities of low-dose RT versus lower rate of subsequent primary emergence).

If an operated patient is found to have high-volume pathologic N2a disease or any higher N stage, postoperative radiation is clearly indicated; for positive margins or extracapsular extension, postoperative chemoradiotherapy is indicated. (See 'Postoperative radiation therapy' below.)

●For patients with clinical stage N2b or higher neck disease, we usually recommend definitive concurrent chemoradiation. Single-modality irradiation of the neck and mucosal axis may be considered for patients with limited reserve who are poor candidates for combined modality treatment. Surgery is reserved for patients with persistent or recurrent disease following radiation-based treatment.

Definitive combined modality therapy consisting of a formal neck dissection and postoperative radiation therapy (with or without concurrent chemotherapy, based on the presence of positive margins or extracapsular extension) is an alternative option, although it is not usually offered for N stage higher than N2b. (See 'Primary surgery' below and 'Postoperative radiation therapy' below and "Locally advanced squamous cell carcinoma of the head and neck: Approaches combining chemotherapy and radiation therapy".)

●RT, with or without concurrent chemotherapy, to putative mucosal sites of primary tumor plays an important role in minimizing the rate of primary emergence of the occult primary. For the large majority of cases, we recommend irradiation targeted at the most likely candidate mucosal sites and both sides of the neck as part of the treatment plan. (See 'Extent of radiation fields' below and 'Radiation dose' below.)

Primary surgery

Treatment approach — Primary surgery as a single-modality therapy for head and neck carcinoma of unknown primary requires curative resection of all gross disease by formal neck dissection. This approach may be sufficient for many patients with pathologically staged N1 neck disease without extracapsular extension (clinical stage T0N1M0) (table 1) [33,34]. However, retrospective data have demonstrated a higher overall locoregional control rate with the addition of postoperative RT, likely due to suppression of subsequent primary site emergence.

On the other hand, patients with more advanced neck disease (as judged by the presence of extracapsular extension or pathologic stage N2 or higher) require postoperative adjuvant RT. Alternatively, if RT is already planned based on clinical staging, definitive RT, with or without chemotherapy, may be initiated, with neck dissection utilized only if needed to treat residual disease persisting after RT. (See "Adjuvant radiation therapy or chemoradiation in the management of head and neck cancer".)

For patients choosing an upfront neck dissection, many centers will conduct the search for an unknown primary while the patient is under anesthesia for surgical management of the neck. For example, one common approach involves removal of the ipsilateral tonsil with intraoperative frozen section, followed by lingual tonsillectomy (if tonsil pathology is negative) and a formal neck dissection [22]. If the primary is identified, resection of the primary site disease will be performed in that same setting.

The risk of recurrence by nodal stage is illustrated by two series:

●A series of 70 patients included five cases with pathologic N1 disease; four of these had no extracapsular extension, and one had only microscopic extracapsular extension [35]. The remaining 65 patients had pN2 or greater disease, and 26 had extracapsular extension. Postoperative RT was recommended for all patients with either extracapsular extension, or N2 or greater disease. None of the N1 patients received adjuvant radiation. The five-year disease-specific survival rate for the entire series was 62 percent. Ipsilateral disease control was achieved in all patients with pN1 disease.

●In another series, 24 patients were initially treated with surgery alone [33]. In this series, six patients (25 percent) had a neck recurrence; five of these had extracapsular extension in their original specimen, and four had stage N2 or higher disease. There were 13 patients with N1 disease, including four with extracapsular extension. The one patient from this group who had relapsed in the neck had extracapsular extension.

Both studies showed a relatively low rate of primary emergence (4 and 11 percent) with treatment restricted to the neck [33,35]. However, other studies have reported a higher rate of primary emergence, especially with surgery alone. A combined analysis of four studies in which surgery alone was the treatment estimated a primary emergence rate of 25 percent if the mucosa at risk was not irradiated [36]. If the mucosa at risk is irradiated, mucosal control is nearly always achieved.

There are no randomized trials comparing postoperative (adjuvant) RT alone with postoperative concurrent chemoradiotherapy in this setting, although emerging evidence in HPV associated oropharyngeal cancers indicates that the contribution of concurrent chemotherapy may be limited in the absence of residual gross disease or pathologically identified macroscopic extracapsular extension. Observational data likewise confirm that benefits of concurrent chemoradiotherapy are limited to patients with more advanced neck disease.

●In one study, 22 of 25 patients (88 percent) with N2 or N3 nodal disease were treated with neck dissection followed by concurrent chemoradiotherapy [37]. With a median follow-up of four years, the five-year overall survival rate was 75 percent, with no emergence of a mucosal primary.

●In another study, 37 patients with predominantly N2 and N3 nodal disease were treated with neck dissection followed by concurrent chemoradiotherapy [38]. At a median follow-up of 42 months, 33 of 37 patients (89 percent) were alive, and the rate of regional recurrence was 5 percent (2 patients out of 37). There was no emergence of a mucosal primary during this time.

These results are consistent with older trials that found that definitive concurrent chemoradiotherapy improves locoregional control, compared with RT alone, for patients with locally advanced SCC of the head and neck. Furthermore, the lack of primary emergence in both of these studies is also important. (See "Locally advanced squamous cell carcinoma of the head and neck: Approaches combining chemotherapy and radiation therapy".)

It is unclear, however, if the increased responsiveness to treatment of HPV associated oropharyngeal cancers may translate into this area. For oropharyngeal cancers, standard treatments in both the definitive and postoperative RT setting are being questioned in several de-escalation trials. (See "Treatment of human papillomavirus associated oropharyngeal cancer", section on 'Is there a role for treatment deintensification?'.)

Moreover, chemoradiotherapy and RT are associated with significant toxicity since both studies had high rates of enteral feeding, hospitalization, and esophageal stricture. (See "Management and prevention of complications during initial treatment of head and neck cancer" and "Management of late complications of head and neck cancer and its treatment".)

Without firm data supporting the routine use of postoperative chemoradiotherapy for unknown primary malignancies of the head and neck and in light of its substantial toxicities, we recommend following standard guidelines restricting the use of chemoradiotherapy to high-risk situations, such as clearly positive margins and/or macroscopic extracapsular extension in the neck.

Postoperative radiation therapy — RT is used to control regional disease in both sides of the neck and prevent the emergence of a primary tumor by treating the putative mucosal sites. In both the postoperative setting and for definitive RT, two important issues that need to be considered are how much of the mucosal surface and neck should be included within the RT field and the minimum radiation dose needed to prevent the emergence of a primary tumor. (See "General principles of radiation therapy for head and neck cancer".)

In older series, RT utilized opposed lateral radiation fields or three-dimensional conformal radiation techniques with the basic framework of an opposed lateral arrangement. These techniques were used to treat the bilateral mucosal surfaces and neck. However, in the contemporary era, techniques such as intensity modulated radiation therapy (IMRT) are now used for most head and neck cancers due to benefits in salivary sparing and quality of life [39-41]. (See "General principles of radiation therapy for head and neck cancer".)

However, the extent of the radiation fields and the optimal radiation dose remain uncertain since many, although not all, patients with SCC of unknown primary may harbor an occult oropharyngeal primary, which has a high likelihood of being HPV associated. For HPV associated oropharyngeal cancer, multiple potentially paradigm-changing clinical trials are in progress at present, testing whether equivalent outcomes may be obtained with reduced radiation doses or reduced radiation fields. (See "Epidemiology, staging, and clinical presentation of human papillomavirus associated head and neck cancer", section on 'Treatment'.)

Definitive radiation therapy — Definitive RT, often including concurrent chemotherapy, is an established treatment for head and neck carcinoma of unknown primary. It is the major alternative approach to an upfront neck dissection with removal of any identified primary site. With definitive RT, neck dissection is reserved for patients who have persistent neck disease at the conclusion of RT. (See "General principles of radiation therapy for head and neck cancer".)

The largest experience with definitive RT comes from the Danish Society of Head and Neck Oncology, which reviewed their experience of 352 patients with head and neck carcinoma of unknown primary over a 20-year period [42]. Twenty-six patients were treated with RT to the ipsilateral neck alone (four of whom had subsequent surgery), 224 (21 of whom then had surgery) were treated with RT to both sides of the neck and to all mucosal sites (the entire pharyngeal axis and larynx), and 23 had surgery alone. Treatment of the bilateral neck and mucosa versus ipsilateral neck alone resulted in a non-statistically significant improvement in neck control (52 versus 43 percent), mucosal control (87 versus 77 percent), locoregional control (48 versus 31 percent), cause-specific survival (47 versus 32 percent), and overall survival (34 versus 22 percent). Patients treated with the ipsilateral neck radiation alone had a relative risk of recurrence in the head and neck region of 1.9 compared with patients treated to both neck and mucosa. The emergence of an occult primary occurred in 20 percent of the entire cohort, but both RT groups were significantly better than surgery in the control of occult mucosal primaries. While there was no statistically significant difference between the two RT groups, the data seemed to indicate superior results with a more inclusive bilateral approach.

This study does confirm that RT controls the emergence of an occult primary. However, contemporary diagnostic technology (ie, positron emission tomography [PET] scanning/computed tomography [CT]) as well as enhanced molecular analysis (p16, HPV, Epstein-Barr virus [EBV] detection) and robotically assisted evaluation have greatly enhanced our ability to identify the primary site.

Several retrospective, observational studies suggest that IMRT may provide effective locoregional disease control while limiting toxicity and potentially improving long-term quality of life compared with older techniques [43-47]. (See "General principles of radiation therapy for head and neck cancer", section on 'Intensity-modulated RT'.)

As an example, 52 patients in one study were treated using IMRT to both sides of the neck and to mucosal sites (pharyngeal axis) [44]; only one patient recurred at a primary site (base of tongue), and three patients recurred in the neck. Long-term complications included one patient requiring percutaneous endoscopic gastrostomy tube feeding and one patient with an esophageal stricture requiring dilation. (See "Management and prevention of complications during initial treatment of head and neck cancer" and "Management of late complications of head and neck cancer and its treatment".)

The addition of chemotherapy to definitive RT is widely used in patients with locoregionally advanced head and neck cancers. By corollary, definitive concurrent chemoradiotherapy is now widely used in high-risk patients with head and neck carcinoma of unknown primary [43-45], such as those with N2-3 neck disease and/or extracapsular extension in the neck. (See "Locally advanced squamous cell carcinoma of the head and neck: Approaches combining chemotherapy and radiation therapy".)

Extent of radiation fields — Assuming that an adequate endoscopic examination under anesthesia and appropriate imaging studies have been conducted, and a primary tumor has not been identified, the most widely accepted current technique is to treat the likely site(s) of primary origin, along with all of the draining lymph node bearing regions of the neck. Sites that are unlikely to contain the primary tumor are often excluded in order to reduce toxicity.

Unless there are lymph nodes in primary drainage basins of the oral cavity, eg, the submental or submandibular regions of the neck, the oral cavity is generally not irradiated. The intraoral mucosa is thin and easily inspected for signs of primary cancer; there is a low likelihood of an occult primary in this location, and the risks of short- and long-term toxicity associated with RT to this region are high.

In patients who present with isolated level 2 adenopathy, especially non-smokers with positive testing for EBV, HPV, or p16, the larynx and hypopharynx are often excluded. In patients with level 3 to 4 adenopathy, especially in the absence of level 2 adenopathy, the larynx and hypopharynx are considered to be likely primary sites and should be irradiated.

Thus, for most patients, a very common treatment volume includes both sides of the neck, the retropharyngeal lymph nodes, and the structures of the oropharynx, with a focus on the soft palate, tonsils, base of tongue, and surrounding pharyngeal walls. Inclusion of the entirety of the nasopharynx is controversial in patients who have undergone appropriate endoscopic and radiographic evaluation, and who do not possess the usual demographic factors associated with nasopharyngeal carcinoma. On the other hand, irradiation of the retropharyngeal lymph node region will often result in some limited coverage of the posterior nasopharynx, and the inclusion of the anterolateral nasopharyngeal mucosa may add relatively little additional toxicity.

Changes to these treatment volume approaches have been proposed, although there are no data directly comparing different radiation field designs. For example, a single-institution experience explored the possibility of restricting radiation only to the oropharynx as the only potential primary site of SCC of unknown primary. The investigators reported a three-year-estimated locoregional control rate of over 95 percent, resulting from treatment directed specifically at the oropharynx along with the bilateral neck for SCC of unknown primary in non-Asian American individuals. Among 68 patients in this series, 82 percent were White American individuals, 75 percent were male, 75 percent had stage IVA cancer, and 56 percent received concurrent chemotherapy; all received a dose of 70 Gy to the oropharynx [48].

The extent of the radiation fields to the neck has also been investigated. Retrospective series suggest that unilateral irradiation may be effective, especially in patients in whom an occult tonsillar primary cancer is suspected on the basis of p16/HPV positivity [49,50]. A randomized study by the European Organisation for Research and Treatment of Cancer (EORTC 22205) conducted in patients with SCC of unknown primary with resected ipsilateral neck disease attempted to evaluate radiation of the ipsilateral neck alone versus radiation to the bilateral neck plus the entire mucosa of the nasopharynx, oropharynx, larynx, and hypopharynx. However, this trial closed early due to poor patient accrual.

Radiation dose — The dose of radiation is based on that used for other SCC of the head and neck region: 70 Gy for gross disease, 60 Gy for the involved, highest risk, or previously involved nodal basins, and 50 Gy to control subclinical or occult disease in the mucosa or potentially draining nodal basins.

Caution should be used in treating an extensive mucosal surface with more than 50 Gy (or 54 Gy in 30 fractions). If large areas of mucosa are irradiated, even moderate doses may be associated with significant toxicity. As an example, a median dose of 60 Gy to all of the pharyngeal mucosa, including the hypopharynx, was used in a series of 24 patients with SCC of unknown primary; in this group, 11 patients (46 percent) required esophageal dilatation for stricture [43]. Thus, if a wide-field, radiation-based approach is used, the highest radiation doses should be reserved for the involved area of the neck (60 to 70 Gy), and mucosal surfaces and node-negative regions of the neck should receive the lowest doses adequate for subclinical disease (50 Gy).

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 AND RECOMMENDATIONS

●Definition – Head and neck carcinoma of unknown primary is defined by the presence of squamous cell carcinoma (SCC) in one or more lymph nodes within the head and neck region that are not solely in the supraclavicular region, without identification of a primary tumor after an exhaustive workup. (See 'Introduction' above.)

●Diagnostic evaluation – The initial diagnostic evaluation of these patients includes a comprehensive history and physical examination, minimally invasive fine needle aspiration of the neck mass, office nasopharyngolaryngoscopy, and a thorough imaging workup (magnetic resonance imaging [MRI], computed tomography [CT], or positron emission tomography [PET]/CT with a high-resolution, contrast-enhanced diagnostic CT component). PET/CT should ideally be performed prior to operative manipulation or invasive biopsies. If the primary tumor is not identified, panendoscopy with bilateral tonsillectomy and directed biopsies is indicated, and transoral lingual tonsillectomy may be considered. (See "Evaluation of a neck mass in adults" and 'Diagnostic evaluation' above.)

●Management – There are two main goals for treatment: control of disease in the neck and prevention of posttreatment primary tumor emergence.

•For patients with N1 disease without extracapsular extension (table 1), we suggest single-modality therapy with either surgery or definitive radiation therapy (RT) (Grade 2C). (See 'Overview' above and 'Primary surgery' above and 'Postoperative radiation therapy' above.)

For patients with clinical stage N2b or higher neck disease, we suggest definitive concurrent chemoradiation (Grade 2C). Single-modality irradiation of the neck and mucosal axis may be considered for patients with limited reserve who are poor candidates for combined modality treatment. Surgery is reserved for patients with persistent or recurrent disease following radiation-based treatment. (See 'Overview' above and 'Definitive radiation therapy' above and "Locally advanced squamous cell carcinoma of the head and neck: Approaches combining chemotherapy and radiation therapy".)

•RT, with or without concurrent chemotherapy, to putative mucosal sites of primary tumor plays an important role in minimizing the rate of primary emergence. As a general standard, we offer radiation that targets the most likely candidate mucosal sites and both sides of the neck. (See 'Overview' above and 'Extent of radiation fields' above.)