INTRODUCTION Cancer of unknown primary site (CUP) is a common clinical entity, accounting for 2 percent of all cancer diagnoses in the Surveillance, Epidemiology, and End Results (SEER) registries between 1973 and 1987 . Within this category, tumors from many primary sites with varying biologies are represented; this heterogeneity has made the design of therapeutic studies difficult.
Substantial improvements have been made in the management and treatment of some patients with CUP. The identification of specific subgroups of treatable patients has been made possible by the development of specialized immunohistologic techniques that can aid in tumor characterization, and by the recognition of several clinical syndromes that permit prediction of chemotherapy responsiveness.
The typical patient with CUP presents with symptoms referable to a metastatic site. The initial work-up, including physical examination, laboratory and radiographic study, fails to identify the primary site. Light microscopic evaluation of biopsy material places the tumor into one of five histologic categories, which then guides further evaluation:
Effective treatment is available for some patients with SCC of unknown primary site who fit certain clinical syndromes, particularly those with involvement of the cervical or inguinal lymph nodes. For this reason, appropriate evaluation of these patients is essential prior to embarking on treatment.
CERVICAL LYMPH NODE INVOLVEMENT The cervical lymph nodes are the most common metastatic site for SCC of unknown primary origin. The probable primary site of origin differs with involvement of upper or lower cervical nodes.
Upper or mid cervical nodes Patients with involvement of the upper or mid cervical nodes are usually middle-aged or elderly, and many have a history of substantial tobacco and/or alcohol use. In such patients, a primary tumor in the head and neck region should be suspected. Optimal initial evaluation includes a thorough examination of the oropharynx, hypopharynx, nasopharynx, larynx, and upper esophagus by direct vision and fiberoptic nasopharyngolaryngoscopy, with biopsy of any suspicious areas. We do not routinely perform bronchoscopy if the chest CT is negative and the patient lacks pulmonary symptoms. In one study, this diagnostic approach identified a primary site in 231 of 267 patients (87 percent) presenting with metastatic SCC in neck nodes.
Computerized tomography (CT) of the neck is useful in defining the extent of disease and occasionally identifies a primary site. Other modalities that have been used to improve diagnostic accuracy include positron emission tomography (PET) scanning, screening tonsillectomy, chromosomal analysis, and detection of Epstein-Barr virus genome in tumor tissue.
Diagnosis Fine needle aspiration biopsy is frequently used to make an initial tissue diagnosis when a patient presents with a metastatic cervical lymph node without an obvious primary site. This technique has high sensitivity and specificity and a diagnostic accuracy that ranges from 89 to 98 percent. Incisional biopsy of a pathologic cervical node should be avoided if possible, since it has been associated with a higher incidence of neck failure and inferior survival following definitive treatment in some but not all studies of patients with head and neck cancer.
PET scanning PET scanning using fluorodeoxyglucose (FDG) may permit the identification of metabolically active, small or superficial lesions that are not seen on CT or MRI. A number of small studies have evaluated PET scanning and integrated PET CT imaging in patients with cervical lymphadenopathy
The potential utility of this approach was illustrated in a series of 42 patients with cervical node metastases from occult SCC, in whom panendoscopy and standard radiographic imaging failed to identify a primary tumor site . Although whole-body FDG-PET imaging showed focal pathologic uptake in 20 (48 percent), additional investigations confirmed a primary tumor in only ten (seven in the head and neck and one each in the lung, esophagus, and abdomen). In these ten patients, the additional information substantially altered treatment planning.
A meta-analysis of 16 studies evaluated the role of PET scanning in 302 patients with cervical lymph node metastases from an unknown primary. The overall sensitivity was 88 percent and specificity was 75 percent; previously unrecognized metastases were identified in 27 percent.
Given these results and the implications for treatment, PET scanning provides useful information frequently enough to be considered part of a standard evaluation.
Surveillance biopsies and karyotype analysis Panendoscopy with systematic blind biopsies of mucosal sites likely to harbor occult tumors (eg, nasopharynx, base of tongue, pyriform sinus, tonsils) is the standard of care in many institutions for patients with metastatic squamous cell cancer in the cervical lymph nodes. Karyotype analysis of the mucosal tissue and the nodal SCC may be especially important if all surveillance site biopsies are histologically negative. In one study, for example, genetic alterations in histologically benign mucosal surveillance biopsies from the upper aerodigestive tract were identical to those found in the metastatic cervical nodes in 10 of 18 patients (55 percent). Three of these patients subsequently developed primary head and neck cancers at the site where the abnormal genetic findings were detected.
Screening tonsillectomy Others advocate ipsilateral tonsillectomy as a diagnostic modality in patients with cervical SCC of unknown primary site, particularly in those with a single node involving the subdigastric, midjugulocarotid, or submandibular areas, or in patients presenting with bilateral subdigastric adenopathy . The possible utility of this approach was illustrated in a study of 87 patients who underwent tonsillectomy as part of the work-up for cervical node metastases presenting as an unknown primary cancer. Overall, 26 percent had a tonsillar primary, and the involved lymph node was subdigastric, submandibular, and midjugulocarotid in 38, 28, and 23 percent, respectively. One of the advantages of documenting an occult tonsillar primary is the ability to avoid irradiation of a normal larynx.
Detection of EBV genome Among head and neck cancers, the Epstein-Barr virus (EBV) genome has been found only in nasopharyngeal carcinoma. As a result, its identification in a metastatic cervical node could lead to the identification of an unsuspected primary site within the nasopharynx . In situ hybridization for EBER-1 (EBV encoded RNA) or polymerase chain reaction (PCR) for EBV genomic DNA can be performed on tissue obtained by fine-needle aspiration biopsy, and should be considered particularly in young patients with poorly differentiated squamous histology in a cervical lymph node.
Treatment When a primary site cannot be identified, patients with SCC of unknown primary site that involves the upper or mid cervical nodes should be treated according to guidelines for locally advanced SCC of the head and neck. In such patients, treatment with definitive radiation to the pharyngeal axis and bilateral necks, radical neck dissection, or a combination of these local modalities has resulted in long-term disease-free survival in 40 to 67 percent of cases . Postoperative radiation may be considered in patients who undergo a radical neck dissection, particularly if extranodal extension or multiple positive nodes are demonstrated.
One potential advantage of radiation compared to surgical management in this setting is that potential primary sites in the head and neck can be included in the radiation field. One study retrospectively compared the outcomes in 85 patients treated with involved field radiation alone to 59 patients treated with radiation to both the nodes and to potential primary sites in the head and neck . Although there was no difference in five-year survival between the two treatment groups, a larger number of primary sites subsequently became evident in the group treated with nodal irradiation only (six versus one).
Bilateral neck irradiation may provide a better outcome than ipsilateral treatment. In one report of 352 patients with squamous cell or undifferentiated cancer involving the cervical lymph nodes without an evident primary site, 277 were managed with bilateral neck irradiation and elective irradiation of sites in the nasopharynx, hypopharynx and larynx, while 26 received ipsilateral nodal irradiation only . The patients treated with ipsilateral neck radiation compared to those receiving bilateral neck radiation had a 1.9-fold higher risk of recurrence in the head or neck (51 versus 27 percent, p = 0.05), and a trend toward lower five year disease-specific survival (28 versus 45 percent).
Combined modality therapy with concurrent chemotherapy and radiation therapy may improve treatment results in patients with locally advanced head and neck cancer. This approach is reasonable in patients with cervical squamous cell cancer and an occult primary site, although only limited experience is available. In one preliminary report of 19 such patients who were treated on a variety of protocols utilizing induction chemotherapy or concomitant chemoradiotherapy, the actuarial five-year overall and disease-free survival rates were 71 and 85 percent, respectively.
Despite definitive treatment, between 10 and 15 percent of affected patients will develop an evident mucosal head or neck primary site within five years, and an equal number will develop distant metastatic disease . Extracapsular extension, multiple positive nodes, and poorly differentiated tumor histology are poor prognostic features, as in patients with known primary sites in the head and neck .
Lower cervical or supraclavicular nodes A primary lung cancer should be suspected when lower cervical or supraclavicular nodes are involved. Fiberoptic bronchoscopy is indicated if the chest radiograph and head and neck examination are unrevealing.
If the fiberoptic bronchoscopy reveals a primary bronchogenic cancer, positive cervical or supraclavicular lymph nodes denote metastatic lung cancer, and patients are treated with standard protocols.
In the absence of a detectable primary site, the results are poor. Nevertheless, patients with no detectable disease below the clavicle should be treated with the same approach as patients with upper cervical nodes, since occasional patients will have long-term disease-free survival.
INGUINAL LYMPH NODE INVOLVEMENT The vast majority (99 percent in one series) of patients with SCC involving inguinal lymph nodes have a detectable primary site in the genital or anorectal area (including the surrounding skin) .
Digital rectal examination and anoscopy should be performed in both sexes to exclude anorectal lesions. Identification of a primary site in these patients is important, since curative therapy is available for carcinomas of the vulva, vagina, cervix, and anus even with regional lymph node involvement (See specific topic reviews). For the occasional patient in whom no primary site is identified, lymphadenectomy with or without postoperative radiation therapy to the inguinal nodal basin sometimes results in long-term survival .
Although experience is limited in this uncommon group of patients, recent success with combined chemotherapy and radiation in other cancers originating in this region (ie, cervix, anus, bladder) suggest a potential role for this approach.
OTHER METASTATIC SITES Metastatic SCC in areas other than the cervical or inguinal lymph nodes usually represents metastasis from a primary lung cancer. CT of the chest and fiberoptic bronchoscopy should be considered if other clinical features suggest the possibility of lung cancer. Chemotherapy with regimens employed in the treatment of non-small cell lung cancer may be considered in patients with good performance status.
Patients with the diagnosis of poorly differentiated SCC should be evaluated carefully, particularly if other clinical features are unusual for lung cancer (eg, young patient, nonsmoker, unusual metastatic sites). This histologic diagnosis is sometimes based upon scant histologic criteria and may include patients with poorly differentiated carcinoma, in whom other diagnoses should be considered. For this reason, additional pathologic evaluation with immunohistochemical stains or electron microscopy should be considered in such patients. If the diagnosis is still unclear, a trial of therapy for poorly differentiated carcinoma should be considered.