The diagnostic evaluation and therapeutic management of a patient with squamous cell carcinoma of an unknown primary (SCCUP) has considerably evolved over recent decades and will likely continue to change as a result of the improving ability to identify small primary tumors and better tailor the implementation of multimodality therapy. By application of the general principles of head and neck oncology, physicians and surgeons are often able to achieve satisfactory control of the disease in patients with SCCUP.
Key points
- •
Enhanced diagnostic techniques have narrowed the population of patients with an unknown primary and should be considered when comparing results to historical literature.
- •
Application of the general principles of head and neck oncology to squamous cell carcinoma of unknown primary generally results in a gratifying control rate.
- •
Evolution of the delivery of radiation therapy requires oncologists to give further consideration to which tissues to target and which to spare.
Introduction
Squamous cell carcinoma of an unknown primary of the head and neck (SCCUP) is defined as metastatic disease in the lymph nodes of the neck without any evidence of a primary tumor of the mucosa after “appropriate investigation.” It is a diagnosis of exclusion: what constitutes an appropriate investigation depends on the expertise of the physician assigning the diagnosis and pretreatment diagnostic investigations.
Epidemiology
SCCUP accounts for approximately 1% to 4% of all cancers of the head and neck. Recent series suggest that an increasing proportion of SCCUP is associated with human papilloma virus (HPV), and that the incidence is increasing at a rate similar to that of “known-primary” oropharynx cancer.
Histology
Most cases of unknown primary of the head and neck are SCC. Adenocarcinoma represents a minority of unknown primary cases and typically presents in the low neck or supraclavicular fossa. A primary presentation of unknown primary adenocarcinoma in this location is suggestive of a primary site below the clavicles. Thyroid cancer, lymphoma, melanoma, and other diseases occasionally present with a palpable neck metastasis, similar to the presentation of SCCUP. Management of these histologies will not be addressed here.
Introduction
Squamous cell carcinoma of an unknown primary of the head and neck (SCCUP) is defined as metastatic disease in the lymph nodes of the neck without any evidence of a primary tumor of the mucosa after “appropriate investigation.” It is a diagnosis of exclusion: what constitutes an appropriate investigation depends on the expertise of the physician assigning the diagnosis and pretreatment diagnostic investigations.
Epidemiology
SCCUP accounts for approximately 1% to 4% of all cancers of the head and neck. Recent series suggest that an increasing proportion of SCCUP is associated with human papilloma virus (HPV), and that the incidence is increasing at a rate similar to that of “known-primary” oropharynx cancer.
Histology
Most cases of unknown primary of the head and neck are SCC. Adenocarcinoma represents a minority of unknown primary cases and typically presents in the low neck or supraclavicular fossa. A primary presentation of unknown primary adenocarcinoma in this location is suggestive of a primary site below the clavicles. Thyroid cancer, lymphoma, melanoma, and other diseases occasionally present with a palpable neck metastasis, similar to the presentation of SCCUP. Management of these histologies will not be addressed here.
Patient evaluation
Mechanisms used to identify a small primary tumor of the head and neck have evolved over the past 70 years. Increasing sophistication of imaging and examination has narrowed the population of patients typically considered as “unknown primary.” This is an important consideration when evaluating SCCUP in the literature; a patient with what was once an “unknown primary” in a historical series may now instead prove to have a small primary tumor.
Clinical Presentation and Imaging
The most common presentation of SCCUP is a painless neck mass that is refractory to antibiotics. In the absence of an identifiable primary, most patients with SCCUP deny upper aerodigestive tract symptoms.
Level II is the most commonly involved with SCCUP. Presentation of SCCUP in level III without concurrent involvement of level II suggests a primary in the supraglottic larynx or hypopharynx, because these locations more commonly drain directly to midneck. Presentation with lymph nodes in the low neck (level IV, supraclavicular fossa) is associated with a primary site below the clavicles. Approximately 50% of cases present with a single lymph node, whereas 10% to 20% may have bilateral involvement.
Three-dimensional anatomic head and neck imaging has consistently identified small primary tumors not appreciated on physical examination. Positron emission tomography (PET) can further help identify small primaries not appreciated by anatomic imaging or physical examination.
A chest radiograph is necessary before initiating therapy to evaluate for distant metastases. More advanced imaging can be obtained at the discretion of the treating physician. Metastases below the clavicle are uncommon, but may be detected incidentally by PET scans undertaken to identify a primary site.
Tissue Confirmation and Tumor Markers
Fine-needle aspiration (FNA) of the presenting lymph node is the preferred method for obtaining a tissue diagnosis. A nondiagnostic FNA with a clinical suspicion of SCCUP should be performed again, perhaps with ultrasound guidance. If necessary to assign diagnosis, open biopsy should be performed only by a surgeon prepared to undertake definitive surgical management (completion neck dissection) as part of the care at that setting. Open biopsy as an isolated procedure to confirm the diagnosis is discouraged because of the possibility for tumor spillage and the disruption of fascial planes that act as a natural barrier to tumor spread.
Testing of the neck nodes for Epstein-Bar virus (EBV) and HPV can focus diagnostic and therapeutic interventions for SCCUP. The presence of EBV detected via in situ hybridization in metastatic lymph nodes has been associated with a nasopharyngeal primary. A positive p16 stain is most commonly associated with HPV-associated oropharynx tumors, although practitioners must be mindful that cancers from other head and neck sites may stain for p16. A considerable number of unknown primary cases reported in modern literature are HPV-associated ( Table 1 ).
| Author, Year | n | Years | p16+, % |
|---|---|---|---|
| Keller et al, 2013 | 35 | 1990–2010 | 74 |
| Demiroz et al, 2014 | 17 | 1994–2008 | 59 |
| Nagel et al, 2014 | 52 | 1996–2011 | 78 86 True unknown primary |
| Desai et al, 2009 | 41 | 2000–2007 | 27 |
| Graboyes et al, 2014 | 71 | 2001–2012 | 92 a |
| Compton et al, 2011 | 25 | 2002–2009 | 28 |
| Durmus et al, 2014 | 22 | 2008–2012 | 95 a |
a Most patients in these 2 series had a small oropharyngeal primary found with transoral techniques and therefore do not represent true SCCUP. The information gained by including them is uncertain.
Panendoscopy and Biopsy
If no potential primary site has been identified by physical examination or if imaging does not disclose a primary site, then an examination under anesthesia (EUA) should be performed. This may include a nasopharyngeal survey; esophagoscopy (especially for a level IV/supraclavicular lymph node); tonsillectomy; and direct laryngoscopy with directed biopsies of the nasopharynx, base of tongue, supraglottic larynx, and piriform sinus. Historically, the yield of panendoscopy yield was high. Although the primary tumor identification at EUA is less frequent in modern series, endoscopy and biopsy are essential parts of the evaluation because of the high rate of false-negative imaging examinations.
In addition to the typical targets for biopsy listed previously, and regardless of imaging findings, directed biopsies should sample all areas of mucosal irregularity or easy bleeding seen on endoscopy.
Role of Lingual or Palatine Tonsillectomy
The palatine tonsils often conceal an “occult” primary, and thus an ipsilateral palatine tonsillectomy has become a standard component of the initial evaluation of SCCUP. Whether an ipsilateral or bilateral tonsillectomy is required is unclear. Although a small but measurable group of patients prove to have a contralateral occult primary, patients spared a contralateral tonsillectomy do not seem to have an increased rate of mucosal relapse or progression.
Recently, advancements in transoral surgery (TOS) have allowed head and neck surgeons to perform a lingual tonsillectomy through the open mouth, frequently demonstrating a T1 tumor ( Table 2 ). It is uncertain whether the extraordinarily high rates of primary identification with TOS will demonstrably affect outcomes.
| Author, Year | n | Years | Prior EUA (Tonsillectomy), % | Prior CT or MRI, % | Prior PET/CT, % | Transoral Surgery Yield (in Either Tonsil or BOT), % |
|---|---|---|---|---|---|---|
| Durmus et al, 2014 | 11 a | 2008–2012 | 100 (0) | 100 | 100 | 63 |
| Nagel et al, 2014 | 36 | 2002–2011 | 100 (0) | — | — | 86 |
| Mehta et al, 2013 | 10 | 2009–2011 | 100 (100) | 100 | 100 | 100 (all in BOT) |
| Graboyes et al, 2014 | 65 | 2001–2012 | 100 (0) | 100 | 65 | 89 |
| Patel et al, 2013 | 18 a | 2010–2013 | 100 (13) | 81 CT 6 MRI | 57 | 72 |
a Reporting in this table limited to the patients without examination/imaging highly suspicious for primary tumor before lingual ± palatine tonsillectomy.
Initial treatment
Most patients with SCCUP are treated with multimodality therapy. Similar to oropharynx cancer, multimodality therapy is most commonly either resection followed by adjuvant radiation (±chemotherapy) or primary chemoradiation (±post-therapy neck dissection). Similar to known-primary mucosal cancer of the head and neck, patients with SCCUP do well with either approach and institutional patterns of care often determine treatment. Efficacy data seem to be similar, so an evaluation of the toxicity of both treatment paradigms is warranted.
Surgery
Primary surgical treatment for SCCUP is a neck dissection. The extent of the neck dissection is determined by the disease presentation. Management with surgery alone for SCCUP requires the accurate identification of patients at low risk for both mucosal emergence and neck failure. Data reporting the outcomes of this management technique are sparse because of the relatively limited group of patients for whom it is appropriate.
The largest series of surgery alone reports the outcomes of 104 patients from 1948 to 1968. During this time, 52 patients were treated with primary radiation (RT) and 28 were treated with surgery plus postoperative RT. In general, the patients treated by surgery alone had a lower disease burden in the neck. “Almost all” patients had an examination with biopsies under general anesthesia and none had cross-sectional imaging. The mucosal failure rate was 20%. The neck failure rate was 13% for patients with Nx/N1 necks and 32% for patients with N2/N3 necks. Other series substantiate that both mucosal progression and contralateral neck failure are uncommon in selected patients, even in the absence of RT ( Table 3 ). The most common sites of recurrence reported for SCCUP, even in the absence of pretreatment imaging, transoral diagnostic techniques, and RT, are in the ipsilateral neck and distantly. The notable exception to this is a Danish experience that demonstrates high rates of mucosal emergence and neck recurrence. Because the stated policy of the Danish health system during the study period was to approach head and neck cancer with primary RT, it seems likely that these results were referable to undetermined selection bias.
| Author, Year | n | Years | NX/N1, % | Prior EUA (w Biopsy), % | Prior CT or MRI, % | Prior PET, % | Mucosal Emergence, % | Ipsilateral Neck Failure, % | Contralateral Neck Failure, % | Distant Failure, % |
|---|---|---|---|---|---|---|---|---|---|---|
| Jesse et al, 1973 | 104 | 1948–1968 | 43 | Almost all (usually) | 0 | 0 | 20 | 24 | 16 | — |
| Coster | 24 | 1965–1987 | 54 | 46 (not specified) | 0 | 0 | 4 | 25 | 8 | 4 |
| Wang | 57 | 1953–1988 | >50 | All (all) | All patients from 1982–1988 | 0 | 11 | 12 | — | 13 |
| Coker | 26 | 1949–1976 | 35 | 69 (not specified) | 0 | 0 | 12 | 8 | 4 | 16 |
| Grau et al, 2000 | 23 | 1975–1995 | 43 | 94 (55) | 30 CT 7 MRI | 1 | 54 | 42 | 42 | — |
| Iganej | 29 | 1969–1994 | 17 | 100 (all after “late 1970s”) | 0 | 0 | 4 | 25 | 8 | 10 |
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
