From the Guidelines
For an unknown primary cancer, I recommend sending a comprehensive panel of tumor markers to help identify the potential origin, including CEA, CA 19-9, CA 125, CA 15-3, AFP, PSA, beta-hCG, LDH, and thyroglobulin, as well as immunohistochemistry staining on tissue samples for CK7, CK20, TTF-1, CDX2, GATA3, PAX8, and ER/PR receptors, as suggested by 1. The diagnostic markers for an unknown primary tumor can be categorized into several groups, including:
- Primary markers, such as CK7 and CK20, which can help identify the potential origin of the tumor, as shown in Table 2 of 1
- Additional markers, such as those for colorectal and Merkel cell carcinoma, which can be used in conjunction with primary markers to narrow down the potential origin, as suggested by 1
- Tumor markers, such as CEA, CA 19-9, and AFP, which can be elevated in certain types of cancer and can help guide the diagnosis, as mentioned in 1
- Immunohistochemistry staining, which can help identify specific tissue origins, such as TTF-1 for lung cancer or CDX2 for gastrointestinal cancer, as recommended by 1
- Genetic and molecular testing, including next-generation sequencing and microsatellite instability testing, which can help identify actionable mutations and guide treatment decisions, as suggested by the example answer.
The combination of these markers, along with clinical presentation and imaging findings, provides the best chance of identifying the primary tumor site, which is crucial for determining appropriate treatment strategies, as emphasized by 1. Some specific examples of how these markers can be used include:
- Elevated AFP suggesting liver or germ cell tumors
- Elevated PSA pointing to prostate cancer
- Specific CK7/CK20 staining patterns distinguishing between gastrointestinal, lung, or gynecological primaries, as shown in Table 2 of 1
- TTF-1 staining suggesting lung cancer
- CDX2 staining suggesting gastrointestinal cancer, as recommended by 1.
It is essential to note that the unique biology of cancers of unknown primary site remains almost unknown, as mentioned in 1, and that chromosomal instability may account for part of the uncommon clinical presentation, chemoresistance, and poor outcome in patients with CUP, as suggested by 1. Therefore, a comprehensive diagnostic work-up, including a thorough physical examination, basic blood and biochemistry survey, urinalysis, fecal occult blood test, and CT scan of thorax, abdomen, and pelvis, as well as whole-body CT/FDG-PET, should be performed to define the extent of tumor dissemination and help identify a minority of CUP patients who can expect to benefit from directed therapy, as recommended by 1.
From the Research
Diagnostic Markers for Cancer of Unknown Primary Origin
The diagnostic markers for cancer of unknown primary origin (CUP) include:
- Histopathologic evaluation 2
- Identification of favorable risk subtypes 2
- Molecular tissue of origin tests to identify the likely primary tumor site 2
- Gene expression profiling 3
- Immunohistochemistry (IHC) studies 3
- Next generation sequencing to identify actionable genomic aberrations 2
- Deep-learning-based algorithms, such as Tumour Origin Assessment via Deep Learning (TOAD), that can predict the origin of the primary tumor using routinely acquired histology slides 4
Diagnostic Challenges
The diagnosis of CUP is challenging due to:
- The absence of a standard-of-care for the initial therapeutic regimen 5
- The impossibility to include these cases in randomized clinical trials 5
- The lack of a identifiable primary tumor site despite a diagnostic work-up 6
- The need for a complex assessment, including a complete medical history, physical examination, complete blood count, urinalysis, serum chemistries, chest radiograph, computed tomography, magnetic resonance imaging, and immunohistochemistry (IHC) studies 3
Current Diagnostic Methods
The current diagnostic methods for CUP include:
- A complete medical history of the patient 3
- Physical examination 3
- Complete blood count 3
- Urinalysis 3
- Serum chemistries 3
- Histologic evaluation 3
- Chest radiograph 3
- Computed tomography 3
- Magnetic resonance imaging 3
- Immunohistochemistry (IHC) studies 3
- Molecular diagnostic information, such as gene expression profiling 3
- Next generation sequencing 2
- Deep-learning-based algorithms, such as TOAD 4