Introduction^*

Cancer of unknown primary (CUP) is a designation given to a discordant group of metastatic carcinomas for which the primary site of origin cannot be identified despite a thorough diagnostic workup that includes a thorough medical history, complete physical examination to include breast and pelvic and rectal evaluations, complete blood count and biochemical analysis, urinalysis, prostate-specific antigen (PSA) in males, and histopathologic review of tissue specimens with immunohistochemistry. Computed tomography (CT) of the chest, abdomen, and pelvis; chest radiography; and mammography in females are performed as indicated.

Despite this workup, the attending physician will more than likely still be in a quandary as to the primary site that produced the aggressive metastatic disease. Needless to say, this scenario creates stress not only to the attending physician but also to the patient in trying to cope with the unknown.

Unproved theories regarding the localization of the primary CUP site include (1) the primary tumor has involuted and only the metastatic disease is evident and (2) metastatic disease is favored over the primary growth based on the phenotype and genotype of the tumor.¹

Epidemiology and Risk Factors

Worldwide, CUP is one of the 10 most frequent cancers constituting about 3% to 5% of all cancer cases.² In 2011, according to the American Cancer Society, there were an estimated 30,500 cases of “Other & unspecified primary sites” in the United States or approximately 2% of all cancer cases and an estimated 44,260 deaths.² At presentation, the median age is approximately 60 years and slightly more frequent in males.³

No risk factors have been identified for this heterogeneous group of neoplasms and no screening programs have been described.

Anatomy and Pathology

Because only carcinomas are included in the diagnosis of CUP, four main histologic types of CUP have been described: well to moderately differentiated adenocarcinomas (50%), undifferentiated or poorly differentiated carcinomas (30%), squamous cell carcinomas (15%), and undifferentiated neoplasms (5%). The latter group includes lymphomas, sarcomas, germ cell tumors, poorly differentiated carcinomas, neuroendocrine tumors, and embryonal malignancies that can be characterized by immunohistologichemistry.⁴ In children, CUPs represent less than 1% of solid tumors and the majority of these tumors are embryonal malignancies.²

Pathologic assessment is a must to characterize a CUP and is usually accomplished by histologic and cytopathologic evaluation. As an adjunct, electron microscopy is occasionally used to evaluate those CUPs that demonstrate indeterminate features that cannot be resolved with routine pathologic evaluation.

Serum tumor markers are mainly overexpressed glycoproteins that are released into the bloodstream by malignant tumors. The serum determination of these tumor markers can help to identify and diagnose, classify, and in patient follow-up, aid in assessment of response to therapy. However, these serum tumor markers have low sensitivity and low specificity, given that they are not expressed specifically by one organ.

The number of serum tumor markers is determined by the attending physician and include but are not limited to carcinoembryonic antigen (CEA), CA125, CA19-9, CDX2, CA15-3, CK-7, CK-20, thyroid transcription factor-1 (TTF-1), PSA, alpha-fetoprotein (AFP), beta-human chorionic gonadotropin (β-HCG), estrogen receptors, and gross cystic disease fluid protein-15 (GCDFP-15). Some of these serum tumor markers can have elevated levels in benign disease such as CEA in inflammatory bowel disease and CA19-9 in pancreatitis.

More recently, molecular techniques have been used to aid in the determination of the site of origin of these CUPs. Tissue microarrays and transcriptional profiling (reverse transcriptase polymerase chain reaction [RT-PCR]) are examples of the available molecular technologies. Some of these molecular assays are commercialized and available by several vendors.

The chromosomal and molecular abnormalities identified in CUPs are beyond the scope of this chapter.

Clinical Presentation

Patients with known primary malignancy have a neoplasm with a natural history and a predictable metastatic pattern that allows for staging. The opposite is true for CUPs, given that they are very aggressive and have an unpredictable metastatic pattern, and therefore, no formal staging can be made in certain patients.

The clinical presentation is due to the symptoms created by the rapid spread of metastases rather than the primary tumor. The presenting symptoms tend to be of short duration, and on physical examination, there may be visual and palpable masses. Constitutional symptoms can also be elicited during the initial evaluation.

These patients face a grim prognosis with reported survival rates of 6 to 10 months for those enrolled in clinical studies.²

However, those CUP patients not enrolled in clinical trials have reported life expectancies of 2 to 3 months.⁵ Given the diverse clinical presentation of these patients, it is best to classify them into favorable and unfavorable subsets, with the former having a better prognosis.⁶

The favorable subsets include

Figure 34-1 A 64-year-old man presented with a “lump” in the left axilla. Left axillary lymph node biopsy shows squamous cell carcinoma. Thorough workup failed to reveal site of origin. A, Coronal positron-emission tomography/computed tomography (PET/CT) scan shows increased activity in the left subpectoral nodes (arrow). B, Coronal PET/CT scan shows increased activity in the left axillary nodes (arrow). C, Axial PET/CT scan shows increased activity in the left subpectoral nodes (arrows).

Figure 34-2 A 54-year-old man presented with a “lump” in the left lower neck. Outside initial biopsy shows poorly differentiated adenocarcinoma. Further immunohistochemistry shows prostate origin. A, Coronal PET/CT shows increased activity in the left lower neck (top arrow), left axillary (middle arrow), and left subpectoral (bottom arrow) adenopathy. B, Coronal PET/CT scan shows increased activity in the retroperitoneum and pelvis (arrow). C, Axial PET/CT scan shows increased activity in the prostate gland (arrow). D, Bone scan shows diffuse metastatic disease.

Figure 34-3 A 65-year-old man presented with right groin mass at an outside institution. Biopsy was consistent with squamous cell carcinoma. Thorough workup failed to reveal primary site of origin. Contrast-enhanced pelvic CT scan shows a large, partially necrotic inguinal nodal mass (arrows).

The unfavorable subsets are

Figure 34-4 A 33-year-old woman presented with cough that did not resolve with conventional therapy. CT study of the chest was done to evaluate the lungs. Lung and left breast biopsies were consistent with poorly differentiated adenocarcinoma not specific for lung or breast primary. A, Contrast-enhanced chest CT scan shows pulmonary metastases (arrows) and a right pleural effusion. B, Contrast-enhanced chest CT scan shows mediastinal (long straight arrow) and hilar (short straight arrow) adenopathy, right pleural effusion (curved arrow), and collapse of the right lower lobe (arrowhead). C, Contrast-enhanced chest CT scan shows a small mass (arrow) in the inner lower quadrant of the left breast. D, Contrast-enhanced abdominal CT scan shows multiple liver metastases (arrows).

To complicate the clinical presentation, more than 50% of CUP patients present with multiple sites of metastatic disease and approximately 30% have three or more organs involved with metastatic disease. In comparison, in patients with known primary tumors, fewer than 15% have metastatic disease in three or more sites.^5,7,8

Patterns of Tumor Spread

The most common sites for CUP metastatic disease are the liver, lungs, bones, and lymph nodes. However, when compared with tumors of known origin, CUPs typically present with unpredictable patterns of spread. For example, when pancreatic and hepatic carcinomas present as CUPs, there is a higher incidence of lung and osseous metastatic disease compared with a known primary originating in the pancreas and liver. Brain and osseous metastatic disease from a CUP originating in the gastrointestinal tract is more commonly seen than that from a known gastrointestinal primary carcinoma. Approximately 50% of pulmonary carcinomas have osseous metastatic disease, whereas CUPs of pulmonary origin have a lower incidence of osseous metastatic disease. Hepatic and pulmonary metastatic disease is seen more frequently in cases that present as a CUP from a prostatic carcinoma than from a known prostatic carcinoma.^9–12