This Article Abstract Full Text (PDF) eLetters: Submit a response to this article Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services E-mail this article link to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints/Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mintzer, D. M. Articles by Greene, G. Search for Related Content PubMed PubMed Citation Articles by Mintzer, D. M. Articles by Greene, G.

Cancer of Unknown Primary: Changing Approaches. A Multidisciplinary Case Presentation from the Joan Karnell Cancer Center of Pennsylvania Hospital

Pennsylvania Hospital, Philadelphia, Pennsylvania, USA

Correspondence: David M. Mintzer, M.D., Joan Karnell Cancer Center, 230 West Washington Square, 2nd Floor, Philadelphia, Pennsylvania 19106, USA. Telephone: 215-829-6088; Fax: 215-829-6104; e-mail: dmmonc{at}aol.com

	LEARNING OBJECTIVES

Top Learning Objectives Abstract Case Presentation Pathology Review and Discussion Molecular Genetic Analysis in... PET Scanning in Unknown... Clinical Approach to the... Patient Follow-up Summary References

 
After completing this course, the reader will be able to:

1. Describe the newer pathologic techniques for defining the site of origin of unknown primary cancers, including immunohistochemistry and molecular genetic techniques.
   
2. List the subsets of patients with unknown primary cancers most likely to achieve long-term survival with appropriate therapies.
   
3. Explain how the newer radiologic techniques such as MRI and PET scanning can help to localize unknown primary cancers.
   

	ABSTRACT

Top Learning Objectives Abstract Case Presentation Pathology Review and Discussion Molecular Genetic Analysis in... PET Scanning in Unknown... Clinical Approach to the... Patient Follow-up Summary References

 
Cancer of unknown primary is a common clinical syndrome, accounting for 2%–5% of cancer patients. A representative case is presented. This heterogenous group of disorders includes entities such as poorly differentiated carcinoma of unknown primary, adenocarcinoma of unknown primary, neuroendocrine carcinoma of unknown primary, squamous cell carcinoma of unknown primary, poorly differentiated (not otherwise specified) cancer of unknown primary, and melanoma of unknown primary. It is crucial to identify those treatment-responsive presentations of unknown primary with the greatest potential for long-term survival.

This discussion emphasizes newer approaches to the diagnosis and treatment of unknown primary cancer, including advances in pathology with immunoperoxidase and molecular genetic techniques, positron emission tomography, and published chemotherapeutic trials. With the increased sophistication of pathologic and radiologic techniques, the frequency of unknown primary cancers will likely continue to decline. Further, as newer and more targeted therapies for specific types of cancer are identified, the previously held nihilism regarding the search for and identification of the primary may become less supportable.

Key Words. Cancer of unknown primary • Cancer of unknown origin • Immunohistochemistry • Molecular genetics • Microarray • PET scan

	CASE PRESENTATION

Top Learning Objectives Abstract Case Presentation Pathology Review and Discussion Molecular Genetic Analysis in... PET Scanning in Unknown... Clinical Approach to the... Patient Follow-up Summary References

 
A 49-year-old female presented with swollen lymph nodes in her left neck. She felt well without any active systemic or localizing symptoms. She had a history of an early cervical cancer treated 15 years previously with cryosurgery and cone resection without known recurrence. She had undergone a benign right breast biopsy 3 years previously. Her past medical history was notable for mitral valve prolapse, tonsillectomy, and appendectomy. She had quit smoking 6 years previously, with a 15 pack-year history. Her family history was positive for colon cancer in her mother. She was on no regular medications and had no allergies.

On physical exam, there were multiple palpable firm nodes measuring 1–2 cm in the lower left cervical and medial left supraclavicular areas. The remainder of her physical examination was negative.

CBC, chemistry panel, and chest x-ray were negative. A biopsy of the left lower cervical nodes revealed poorly differentiated carcinoma. Further workup included a computerized tomography (CT) scan of the chest, abdomen, and pelvis, which confirmed the left supraclavicular adenopathy and also showed adenopathy in the upper mediastinum. Ear, nose, and throat evaluation revealed no evident primary lesion on fiberoptic endoscopy. Mammograms were negative. Breast magnetic resonance imaging (MRI) showed no suspicious lesions. Pelvic examination showed no evidence of pathology. Upper and lower endoscopies of the gastrointestinal (GI) tract were negative. Positron emission tomography (PET) scanning revealed uptake in the lower left neck, upper mediastinum, and right hilum, but nowhere else.

	PATHOLOGY REVIEW AND DISCUSSION

Top Learning Objectives Abstract Case Presentation Pathology Review and Discussion Molecular Genetic Analysis in... PET Scanning in Unknown... Clinical Approach to the... Patient Follow-up Summary References

 
Michael Warhol, M.D.
A left scalene node biopsy was received. Within the lymph node there were punched out areas representing metastatic tumor (Fig. 1). Higher powered imaging revealed undifferentiated cells that were not forming any recognizable structures (Fig. 2). So while they were not telling us what they were, even undifferentiated cells can be characterized as to size and shape. An internal measuring stick for size is a lymphocyte. These particular tumor cells were many times bigger than lymphocytes; hence, they were large cells. Their shape was round; in particular, their nuclei were round to oval and had at least one, and in certain cases many, prominent nucleoli. So this was a large round cell tumor. Large round cell tumors are usually either undifferentiated carcinoma, large cell lymphoma, or amelanotic melanoma.

View larger version (162K): [in this window] [in a new window]   Figure 1. Photomicrograph of the scalene node. The lymph node architecture is disrupted by punched out areas of pale-staining metastatic carcinoma (hematoxylin and eosin staining, magnification = 40x).

View larger version (151K): [in this window] [in a new window]   Figure 2. A higher power view of the metastatic tumor. The tumor cells have large, round, vesicular nuclei with prominent nucleoli. There is abundant pale-staining cytoplasm. Cell membranes are distinct (hematoxylin and eosin staining, magnification = 200x).

Immunohistochemistry (IHC) is the current method of choice for the classification of undifferentiated tumors [1]. Different antigens characterize different cell types, with certain antigens present in epithelial tumors, others in lymphoid tumors, and still others in melanomas. The antigens used are specific for only one cell type. Antibodies are colorless, so to identify the antibody, a detection system such as peroxidase enzymes is necessary. These reduce a substrate and produce a brown color at the site of antibody localization.

Intermediate filaments, which form part of the cell cytoskeleton, are particularly useful antigens in identifying cells of origin. Keratins are a family of intermediate filaments that characterize epithelial tissue. The tumor in this case stained strongly with antikeratin antibodies, indicating that it was a carcinoma. Ideally, one would like organ-specific antigens, that is, antigens that are present only in a particular organ. Such antigens do exist; prostate-specific antigen and prostatic acid phosphatase are found only in the prostate. They have a sensitivity of about 95% and are very effective diagnostic tools. Obviously, prostate antigens are not relevant in this case. The gross cystic disease fluid protein or Brst-1 antigen is reasonably characteristic of breast cancer [2]. This antigen, however, only has a sensitivity of about 50%, so it is not expressed in about half of all breast cancers. Hormone receptors, such as estrogen and progesterone receptors, can also be relatively specific for tumors like breast and endometrial cancer, but they also can be found in smooth muscle.

The keratin proteins, a family of approximately 23 different polypeptides, are emerging as relatively organ-specific antigens. They are both acidic and basic polypeptides, varying in molecular weight from 40–67 kDa. In vivo, keratin peptides exist in pairs. Two basic polypeptides match with two acidic polypeptides to form a keratin tetramer. These different pairs are given different numbers. The distribution of keratins varies among the different types of epithelium. Squamous epithelium contains different keratins from glandular epithelium.

Particularly useful keratins include CK7 and CK20 [3, 4] (Table 1). The tumor in this patient was positive for CK7 (Fig. 3) and negative for CK20. So, as shown in Table 1, in the context of the relevant clinical information, ovarian carcinoma and mesothelioma were not considerations. Endometrial carcinoma was not a consideration clinically and neither was pancreatic carcinoma. So we were left with breast cancer and lung cancer as likely possibilities. Now, in addition to being CK7⁺ and CK20^–, the tumor was also carcinoembryonic antigen (CEA) positive. CEA is relatively nonspecific, but it is present frequently in colorectal, non-small cell lung, hepatocellular, and mucinous carcinomas (Table 2). In this particular case, the negative CK20 ruled out colorectal carcinoma. Hepatocellular carcinoma was not a clinical consideration and neither was mucinous carcinoma. So the combination of CK7 positivity and CEA positivity indicates that the tumor in this case was consistent with a lung primary.

View larger version (155K): [in this window] [in a new window]   Figure 3. The scalene node stained with antikeratin 7, using the immunoperoxidase technique. Note the strong diffuse staining of the tumor cells (immunoperoxidase stain, magnification = 100x).

It should be emphasized that the results of IHC staining are but one piece of the puzzle, and that these results must be interpreted in the appropriate context. Although extremely helpful, tissue antigens such as CK7 and CK20 themselves rarely specifically identify a primary site. But, as with this case, they can help suggest a primary site through the process of elimination. Also important to remember is that, by definition, tumors are abnormal and, therefore, may not completely mimic the phenotype of their corresponding normal tissue.

	MOLECULAR GENETIC ANALYSIS IN UNKNOWN PRIMARY CANCER

Top Learning Objectives Abstract Case Presentation Pathology Review and Discussion Molecular Genetic Analysis in... PET Scanning in Unknown... Clinical Approach to the... Patient Follow-up Summary References

 
A.M. Martin, Ph.D.
While molecular genetic analysis was not performed in the case presented here, I review some published results and the potential utility of such analyses in the classification of cancers of unknown primary.

Pathologists classify tumors according to the sites in which they arise and by their morphologies; however, this method of classification can be imprecise. As reviewed above, further analysis relies on the use of IHC techniques to determine the expression of specific molecular markers (proteins). However, even with IHC techniques, limitations exist due to subjectivity, interobserver variation, and the paucity of specific and reliable measurable molecular markers. Standard cytogenetic techniques have also been reported to be occasionally helpful in the classification of unknown primary cancer [5, 6].

Recent technologic advances allow for the simultaneous analysis of genes and their products. Previously, scientists had only been able to conduct specific genetic studies, one gene at a time. However, with the publication of the human genome sequence and the development of DNA microarray technology, we can now study thousands of genes at one time. This greatly enhances the capacity to study every gene in the human genome, identifying and quantifying changes that occur during the course of tumorigenesis.

DNA arrays consist of thousands of short strands of DNA sequences (oligonucleotides), identical to the sequences of normal genes, which are assembled systematically onto the surface of a glass slide [7]. These slides can hold more than 20,000 different oligonucleotides representing more than 20,000 different genes. RNA isolated from the test sample (tumor) and a reference sample (normal adjacent tissue) are reverse-transcribed to cDNA, then each sample is labeled with a different fluorescent dye. The fluorescent samples are hybridized to the DNA array, and the resulting fluorescence intensities indicate gene activity (Fig. 4). Finally, mathematical algorithms are employed to arrange (cluster) differences in gene expression between different samples.

View larger version (23K): [in this window] [in a new window]   Figure 4. DNA array technology. Tumor cDNA and normal cDNA are labeled with different fluorescent dyes. The labeled cDNAs are hybridized to DNA arrays containing oligonucleotides representing known genes. The fluorescence intensity is analyzed using statistical algorithms, and color intensities are assigned to the ratios of gene expression.

The use of DNA microarray technologies has been demonstrated to identify genetic characteristics of various adenocarcinomas [8–12], and the results of these studies are helpful in cases where the adenocarcinoma is of unknown origin. Giordano et al. studied gene expression profiles of 154 primary adenocarcinomas of the lung, colon, and ovary [13]. Using high-density oligonucleotide arrays with 7,129 gene probe sets, they demonstrated comprehensive expression profiles of 57 lung, 51 colon, and 46 ovarian adenocarcinomas. Following statistical analysis, 152/154 of the adenocarcinomas could be classified in an organ-specific manner. Su et al. also reported the use of gene expression signatures in the molecular classification of carcinomas, finding gene subsets that predicted the site of origin for 90% of 175 carcinomas [14]. Finally, Dennis et al. used hierarchical clustering of public expression data from serial analyses of gene expression to identify candidate tumor markers with characteristic patterns and genes predictive for individual primary sites [15].

These findings are both interesting and encouraging, as they confirm what pathologists already suspected as far as histopathological similarities and differences in epithelial-derived tumors are concerned. Furthermore, these studies demonstrate that gene-expression profiles are useful in the reclassification of tumors and can define organ-specific gene expression profiles. Some caution may be needed, however, before assuming that cancers of unknown primary, with their atypical clinical presentations, will have the same molecular genetics of cancers of known primary origin.

Despite these advances, it is unlikely that gene expression analysis will become the gold standard for pathological diagnostic purposes, but instead, it will provide a means for identifying new tumor-specific markers that can be identified using IHC in formalin-fixed paraffin-embedded tissue. To this end, the goal for pathology diagnostics is to use these new technologies to identify tumor-specific protein panels that can be classified based on their expression patterns. Then, tumors can be further characterized and compared simultaneously with a large number of similar tumor samples using tissue microarrays and IHC techniques, ultimately subclassifying them according to their molecular genetic and protein profiles.

In summary, the combination of DNA and tissue-array technologies will greatly facilitate the ability of pathologists to conduct high-throughput validation of tumors. This ultimately will provide a more comprehensive approach to tumor classification and may result in more specific treatment regimens for patients.

	PET SCANNING IN UNKNOWN PRIMARY CANCER

Top Learning Objectives Abstract Case Presentation Pathology Review and Discussion Molecular Genetic Analysis in... PET Scanning in Unknown... Clinical Approach to the... Patient Follow-up Summary References

 
Gary Greene, M.D.
PET is a noninvasive nuclear medicine examination used primarily for the detection of tumors. The patient is injected with a glucose analogue that is radioactive (¹⁸F fluoro-D-glucose [FDG]). This radiotracer travels preferentially to malignant tumors, most of which show increased glucose metabolism. The ¹⁸F FDG emits a positron within the tumor mass, which then combines with an electron, causing emission of two 511 kev gamma photons. These "annihilation" photons travel 180 degrees apart from each other. A positron camera is used to detect the photons and construct tomographic cross-sectional images, similar to CT and MRI, which show up as focal hot areas, similar to other nuclear medicine exams.

While oncologic PET imaging is approved for the detection and/or staging of lung, breast, esophageal, colorectal, and head and neck cancers, and lymphoma, melanoma, and the solitary pulmonary nodule, several series have reported results in the search for the unknown primary. Recent series have reported the utility of PET imaging in patients specifically with cervical lymph node presentations [16–22] as well as in other sites [23–26]. Those studies indicate that PET imaging can detect an unknown primary tumor in 20%–60% of patients when conventional workups have failed to do so. Further experience is still needed to assess how informative and useful PET scans can be in the clinical management of unknown primary patients and whether PET should be considered in all or just selected patients.

The particular patient discussed here presented with left scalene nodes and a CT scan that was positive for pretracheal and precarinal adenopathy. The subsequent PET scan also clearly showed right paratracheal uptake and right hilar uptake as well as confirming disease in the neck within the posterior cervical chain and the left supraclavicular region. In this case the results were consistent with but not diagnostic of a lung primary.

	CLINICAL APPROACH TO THE PATIENT AND THERAPY

Top Learning Objectives Abstract Case Presentation Pathology Review and Discussion Molecular Genetic Analysis in... PET Scanning in Unknown... Clinical Approach to the... Patient Follow-up Summary References

 
David M. Mintzer, M.D.
The literature is replete with excellent reviews on the approach to the patient with cancer of unknown primary [27–31]. It cannot be overemphasized that the initial approach is a careful history, physical exam, and review of the pathology with a workup directed toward sites suggested by this initial evaluation. What may be debated is where and how far to go after that.

Early studies emphasized the futility, expense, and delay caused by extensive workups in search for the unknown primary. In the past, the type of extensive workup often decried might have involved a chest x-ray, sigmoidoscopy, intravenous pyelogram, barium enema, and upper GI. In a woman, a pelvic examination and a mammogram would have been added. This workup might have taken place over 1–2 weeks, would frequently have been performed in the hospital, and would have had a low diagnostic yield of little clinical relevance with a high rate of false positivity. In addition, there was the discomfort, expense, and inconvenience of a prolonged workup to patients who, overall, had lifespans of generally just a few months with limited therapeutic options [27–28].

Presently, however, imaging studies are faster, more accurate, and more available, so that a CT scan of the chest, abdomen, and pelvis can be completed in just a few minutes. In searching for an upper or lower GI primary, a colonoscopy and upper endoscopy can be performed on the same day with a single sedation for patients who have, for example, liver metastases, ascites, or other infra-diaphragmatic presentations, even in the absence of other GI symptoms or signs. And while some would still note that these workups add unnecessary expense, in comparison with the cost of a therapeutic regimen, the cost of these studies, done expeditiously on an outpatient basis, is relatively modest.

Thus, given the improvements in pathologic analysis discussed above, in radiologic imaging (including PET imaging), and in endoscopic approaches, we can now identify the primary site more frequently than in the past. Indeed, our own Pennsylvania Hospital Tumor Registry results reveal a trend indicating that the percentage of unknown primary cancers has declined over the past decades to only 2%, with a similar percentage from the Hospital of the University of Pennsylvania Cancer Center registry.

Furthermore, not only are we now able to find the primary more frequently, but it is also now more relevant therapeutically to know the primary. In the past, if chemotherapy was to be considered, an empiric regimen was chosen. However, with the availability of many newer and more relatively specific and (at least somewhat) more effective chemotherapeutic agents and regimens, it has become increasingly important to define the primary site. For example, in regard to GI adenocarcinomas, there has been a more targeted therapeutic approach as newer agents have become available. Until the past decade, 5-fluorouracil (5-FU)-based therapy was the mainstay for essentially all these tumors, be they colonic, pancreatic, or gastroesophageal in origin. However, at the present time, each of these entities might be treated with a more individualized approach (e.g., gemcitabine for pancreatic, 5-FU/leucovorin with irinotecan or oxaliplatin for colon, and perhaps epirubicin, cisplatin, and 5-FU or a taxane for gastroesophageal primaries). These changes alter the diagnostic and therapeutic nihilism of the past in searching for the unknown primary (Table 3).

Worth emphasizing are some specific unknown primary scenarios to which we have changed our approach and with which long-term survival is sometimes achievable (Table 4). Unknown primary squamous cell carcinoma of the cervical lymph nodes, while less common than in the past, is still seen with some frequency [33, 34]. Given better imaging with CT and MRI scans and with fiberoptic endoscopy (including panendoscopy that encompasses head and neck endoscopy, esophagoscopy, and bronchoscopy) with "blind" biopsies of sites such as the tongue base, tonsil, and nasopharynx, the primary site can be located in many cases. Metastatic disease presenting in level II nodes is commonly found to be tonsillar in origin, and ipsilateral tonsillectomies have been reported to yield the primary site in a significant portion of these cases [35, 36]. Further, molecular genetic studies have shown identical genotyping in areas that appear normal endoscopically but harbor the clonal origins of some of these occult lesions [37]. Even when the primary remains unknown, curative treatment is possible with radiation therapy, surgery, or a combined approach. The sequencing of these modalities, the size of the radiation field, and the role of chemotherapy are controversial and may depend on the extent of the disease and the institutional approach. Nonetheless, up to 50% of these patients may be long-term survivors [33, 34].

The occult breast primary is also diminishing, as mammographic resolution improves and is coupled with ultrasonography and MRI scanning. A breast cancer is usually presumed when a women presents with adenocarcinoma in the axillary nodes without an evident primary [39–44]. MRI has been particularly helpful when there is a high suspicion of a breast primary with a negative physical exam and mammogram [43, 44]. In women presenting with adenocarcinoma in axillary nodes where still no primary can be isolated, an alternative to the past approach of mastectomy has been increasingly used. This new approach involves radiation therapy to the ipsilateral breast and axilla, coupled with whatever appropriate adjuvant systemic therapy would be given for breast cancer based on stage, age, and hormonal status [39–44].

Women, and occasionally men, who present with ascites and diffuse peritoneal seeding may have primary peritoneal cancer [45–47]. These patients generally require laparotomy for diagnosis and may therapeutically benefit from debulking. In these cases, the ovaries have sometimes been removed previously (with documented benign histology) or, if present, are either uninvolved with cancer or have only surface (metastatic) involvement. Clinically and histologically, such peritoneal cancers behave like ovarian cancer with a potential for remission and occasional long-term survival with platinum-based chemotherapy. These cases need to be distinguished from primary peritoneal mesothelioma.

Some unknown primary carcinomas may show neuroendocrine features by IHC or electron microscopy [48]. Those that are poorly differentiated may be more likely to respond to small cell lung cancer-type regimens (e.g., etoposide and platinums).

The scenario of cancer of unknown primary where a lesion was recognized as cancer but not otherwise discernible as carcinoma, lymphoma, sarcoma, or melanoma is all but gone. While a few cases may still defy categorization, the use of immunoperoxidase staining for keratins, vimentin, and common leukocyte antigen, as well as the availability of molecular genetic assays for immunoglobulin and T-cell receptor rearrangements has allowed categorization of most of these entities to at least a broad group of tumors, helping to focus workup and therapy.

The unrecognized midline extragonadal germ cell tumor presentation is critically important to consider, as these patients, generally young males, can be cured with platinum-based testicular cancer chemotherapy programs [49]. They typically present with bulky disease in the retroperitoneum or mediastinum. Measurement of and staining for -fetoprotein and ß-human chorionic gonadotropin in serum and tissue may be helpful. Evidence of karyotypic abnormalities involving chromosome 12 are suggestive of germ cell tumor [5].

When no primary site or particular metastatic pattern is identified, one must resort to an empiric chemotherapy regimen in patients appropriate for treatment. A number of phase II and some phase III trials have been reported [30, 49–52]. However, their interpretation and comparability is problematic since the patients in those trials, by definition, had a variety of different, albeit unknown, primary sites. Nonetheless, a number of investigators, including particularly Hainsworth and Greco [30, 48, 52, 54, 55], have reported results. More recent studies have evaluated the use of a taxane/carboplatin [53] regimen coupled with etoposide [54] or gemcitabine [55].

	PATIENT FOLLOW-UP

Top Learning Objectives Abstract Case Presentation Pathology Review and Discussion Molecular Genetic Analysis in... PET Scanning in Unknown... Clinical Approach to the... Patient Follow-up Summary References

 
The patient described in this report was treated with paclitaxel and carboplatin for three cycles, a regimen with potential activity against a wide variety of primaries. Based on her clinical presentation with supradiaphragmatic disease, her smoking history, and the pathology, we remained suspicious of a lung primary but were never able to prove it. On follow-up, no primary within the lung ever became evident. After a minor response to chemotherapy alone, radiation was added (to encompass all known sites of disease) along with low-dose weekly chemotherapy. She had a partial response but ultimately developed progressive brain metastases poorly responsive to radiation and died 8 months after her initial diagnosis.

	SUMMARY

Top Learning Objectives Abstract Case Presentation Pathology Review and Discussion Molecular Genetic Analysis in... PET Scanning in Unknown... Clinical Approach to the... Patient Follow-up Summary References

 
With this case presentation, we have reviewed cancer of unknown primary with an emphasis on newer concepts and approaches. These include more refined pathologic analyses with IHC and molecular genetics to help identify the primary, and the use of PET scanning, breast MRI, and fiberoptic endoscopy. As a result of these advances in diagnostic approaches, the frequency of cancers remaining of unknown origin is diminishing, although not disappearing. Certain presentations with the greatest potential for long-term survival (axillary node adenocarcinoma, peritoneal carcinomatosis, squamous cervical lymph node metastasis, poorly differentiated neuroendocrine carcinoma, unrecognized germ cell tumors) are crucial to identify. However, even for other presentations, as antineoplastic therapy becomes more specific and effective for given primary sites, it becomes increasingly relevant to identify the site of origin. This represents a further change from the nihilism of the past, both in trying to establish the site of origin and in the treatment of these patients.

	REFERENCES

Top Learning Objectives Abstract Case Presentation Pathology Review and Discussion Molecular Genetic Analysis in... PET Scanning in Unknown... Clinical Approach to the... Patient Follow-up Summary References

 

1. DeYoung BR, Wick MR. Immunohistologic evaluation of metastatic carcinomas of unknown origin: an algorithmic approach. Semin Diagn Pathol 2000;17:184–193.[Medline]
2. Mazoujian G, Warhol MJ, Haagensen DE Jr. The ultrastructural localization of gross cystic disease fluid protein (GCDFP-15) in breast epithelium. Am J Pathol 1984;116:305–310.[Abstract]
3. Wang NP, Zee S, Zarbo RJ et al. Coordinate expression of cytokeratins 7 and 20 define unique subsets of carcinoma. Appl Immunohistochem Mol Morphol 1995;3:99–107.
4. Chu P, Wu E, Weiss LM. Cytokeratin 7 and cytokeratin 20 expression in epithelial neoplasms: a survey of 435 cases. Mod Pathol 2000;13:962–972.[CrossRef][Medline]
5. Motzer RJ, Rodriguez E, Reuter VE et al. Molecular and cytogenetic studies in the diagnosis of patients with poorly differentiated carcinomas of unknown primary site. J Clin Oncol 1995;13:274–282.[Abstract/Free Full Text]
6. Pantou D, Tsarouha H, Papadopoulou A et al. Cytogenetic profile of unknown primary tumors: clues for their pathogenesis and clinical management. Neoplasia 2003;5:23–31.[Medline]
7. Lockhart DJ, Dong H, Byrne MC et al. Expression monitoring by hybridization to high-density oligonucleotide arrays. Nat Biotechnol 1996;14:1675–1680.[CrossRef][Medline]
8. Boer JM, Huber WK, Sultmann H et al. Identification and classification of differentially expressed genes in renal cell carcinoma by expression profiling on a global human 31,500-element cDNA array. Genome Res 2001;11:1861–1870.[Abstract/Free Full Text]
9. Chakrabarti R, Robles LD, Gibson J et al. Profiling of differential expression of messenger RNA in normal, benign, and metastatic prostate cell lines. Cancer Genet Cytogenet 2002;139:115–125.[CrossRef][Medline]
10. Virtanen C, Ishikawa Y, Honjoh D et al. Integrated classification of lung tumors and cell lines by expression profiling. Proc Natl Acad Sci USA 2002;99:12357–12362.[Abstract/Free Full Text]
11. Hasegawa S, Furukawa Y, Li M et al. Genome-wide analysis of gene expression in intestinal-type gastric cancers using a complementary DNA microarray representing 23,040 genes. Cancer Res 2002;62:7012–7017.[Abstract/Free Full Text]
12. Sawiris GP, Sherman-Baust CA, Becker KG et al. Development of a highly specialized cDNA array for the study and diagnosis of epithelial ovarian cancer. Cancer Res 2002;62:2923–2928.[Abstract/Free Full Text]
13. Giordano TJ, Shedden KA, Schwartz DR et al. Organ-specific molecular classification of primary lung, colon, and ovarian adenocarcinomas using gene expression profiles. Am J Pathol 2001;159:1231–1238.[Abstract/Free Full Text]
14. Su AI, Welsh JB, Sapinoso LM et al. Molecular classification of human carcinomas by use of gene expression signatures. Cancer Res 2001;61:7388–7393.[Abstract/Free Full Text]
15. Dennis JL, Vass JK, Wit EC et al. Identification from public data of molecular markers of adenocarcinoma characteristic of the site of origin. Cancer Res 2002;62:5999–6005.[Abstract/Free Full Text]
16. Braams JW, Pruim J, Kole AC et al. Detection of unknown primary head and neck tumors by positron emission tomography. Int J Oral Maxillofac Surg 1997;26:112–115.[Medline]
17. Stokkel MP, Terhaard CH, Hordijk GJ et al. The detection of unknown primary tumors in patients with cervical metastases by dual-head positron emission tomography. Oral Oncol 1999;35:390–394.[CrossRef][Medline]
18. Aassar OS, Fischbein NJ, Caputo GR et al. Metastatic head and neck cancer: role and usefulness of FDG PET in locating occult primary tumors. Radiology 1999;210:177–181.[Abstract/Free Full Text]
19. Safa AA, Tran LM, Rege S et al. The role of positron emission tomography in occult primary head and neck cancers. Cancer J Sci Am 1999;5:214–218.[Medline]
20. Jungehulsing M, Scheidhauer K, Damm M et al. 2[F]-fluoro-2-deoxy-D-glucose positron emission tomography is a sensitive tool for the detection of occult primary cancer (carcinoma of unknown primary syndrome) with head and neck lymph node manifestation. Otolaryngol Head Neck Surg 2000;123:294–301.[CrossRef][Medline]
21. Dede F, Ajoedi ND, Ansari SM et al. Metastatic thyroid cancer occurring as an unknown primary lesion: the role of F-18 FDG positron emission tomography. Clin Nucl Med 2001;26:396–399.[CrossRef][Medline]
22. Nieder C, Gregoire V, Ang KK. Cervical lymph node metastases from occult squamous cell carcinoma: cut down a tree to get an apple? Int J Radiat Oncol Biol Phys 2001;50:727–733.[CrossRef][Medline]
23. Kole AC, Nieweg OE, Pruim J et al. Detection of unknown occult primary tumors using positron emission tomography. Cancer 1998;82:1160–1166.[CrossRef][Medline]
24. Lassen U, Daugaard G, Eigtved A et al. ¹⁸F-FDG whole body positron emission tomography (PET) in patients with unknown primary tumors (UPT). Eur J Cancer 1999;35:1076–1082.
25. Bohuslavizki KH, Klutmann S, Kroger S et al. FDG PET detection of unknown primary tumors. J Nucl Med 2000;41:816–822.[Abstract/Free Full Text]
26. Rades D, Kuhnel G, Wildfang I et al. Localised disease in cancer of unknown primary (CUP): the value of positron emission tomography (PET) for individual therapeutic management. Ann Oncol 2001;12:1605–1609.[Abstract/Free Full Text]
27. Nystrom JS, Weiner JM, Heffelfinger-Juttner J et al. Metastatic and histologic presentations in unknown primary cancer. Semin Oncol 1977;4:53–58.[Medline]
28. Stewart JF, Tattersall MH, Woods RL et al. Unknown primary adenocarcinoma: incidence of overinvestigation and natural history. Br Med J 1979;1:1530–1533.
29. Abbruzzese JL, Abbruzzese MC, Lenzi R et al. Analysis of a diagnostic strategy for patients with suspected tumors of unknown origin. J Clin Oncol 1995;13:2094–2103.[Abstract/Free Full Text]
30. Hainsworth JD, Greco FA. Management of patients with cancer of unknown primary site. Oncology (Hunting) 2000;14:563–574; discussion 574–576, 578–579.
31. Hess KR, Abbruzzese MC, Lenzi R et al. Classification and regression tree analysis of 1,000 consecutive patients with unknown primary carcinoma. Clin Cancer Res 2000;5:3403–3410.
32. Schlagenhauff B, Stroebel W, Ellwanger U et al. Metastatic melanoma of unknown primary origin shows prognostic similarities to regional metastatic melanoma: recommendations for initial staging examinations. Cancer 1997;80:60–65.[CrossRef][Medline]
33. de Braud F, Heilbrun LK, Ahmed K et al. Metastatic squamous cell carcinoma of an unknown primary localized to the neck. Advantages of an aggressive treatment. Cancer 1989;64:510–515.[CrossRef][Medline]
34. Mendenhall WM, Mancuso AA, Amdur RJ et al. Squamous cell carcinoma metastatic to the neck from an unknown head and neck primary site. Am J Otolaryngol 2001;22:261–267.[CrossRef][Medline]
35. McQuone SJ, Eisele DW, Lee DJ et al. Occult tonsillar carcinoma in the unknown primary. Laryngoscope 1998;108:1605–1610.[CrossRef][Medline]
36. Randall DA, Johnstone PA, Foss RD et al. Tonsillectomy in diagnosis of the unknown primary tumor of the head and neck. Otolaryngol Head Neck Surg 2000;122:52–55.[CrossRef][Medline]
37. Califano J, Westra WH, Koch W et al. Unknown primary head and neck squamous cell carcinoma: molecular identification of the site of origin. J Natl Cancer Inst 1999;91:599–604.[Abstract/Free Full Text]
38. Guarischi A, Keane TJ, Elhakim T. Metastatic inguinal nodes from an unknown primary neoplasm. A review of 56 cases. Cancer 1987;59:572–577.[CrossRef][Medline]
39. Kemeny MM, Rivera DE, Terz JJ et al. Occult primary adenocarcinoma with axillary metastases. Am J Surg 1986;152:43–47.[CrossRef][Medline]
40. Rosen PP, Kimmel M. Occult breast carcinoma presenting with axillary lymph node metastases: a follow-up study of 48 patients. Hum Pathol 1990;21:518–523.[CrossRef][Medline]
41. Ellerbroek N, Holmes F, Singletary E et al. Treatment of patients with isolated axillary nodal metastases from an occult primary carcinoma consistent with breast origin. Cancer 1990;66:1461–1467.[CrossRef][Medline]
42. Baron PL, Moore MP, Kinne DW et al. Occult breast cancer presenting with axillary metastases. Updated management. Arch Surg 1990;125;210–214.[Abstract/Free Full Text]
43. Orel SG, Weinstein SP, Schnall MD et al. Breast MR imaging in patients with axillary node metastases and unknown primary malignancy. Radiology 1999;212:543–549.[Abstract/Free Full Text]
44. Olson JA Jr, Morris EA, Van Zee KJ et al. Magnetic resonance imaging facilitates breast conservation for occult breast cancer. Ann Surg Oncol 2000;7:411–415.[CrossRef][Medline]
45. Chen KT, Flam MS. Peritoneal papillary serous carcinoma with long-term survival. Cancer 1986;58:1371–1373.[CrossRef][Medline]
46. Strnad CM, Grosh WW, Baxter J et al. Peritoneal carcinomatosis of unknown primary site in women: a distinctive subset of adenocarcinoma. Ann Intern Med 1989;11:213–217.
47. Dalrymple JC, Bannatyne P, Russell P et al. Extraovarian peritoneal serous papillary carcinoma: a clinicopathologic study of 31 cases. Cancer 1989;64:110–115.[CrossRef][Medline]
48. Hainsworth JD, Johnson DH, Greco FA. Poorly differentiated neuroendocrine carcinoma of unknown primary site: a newly recognized clinicopathologic entity. Ann Intern Med 1988;109:364–371.
49. Hainsworth JD, Greco FA. Extragonadal germ cell tumors and unrecognized germ cell tumors. Semin Oncol 1992;19:119–127.[Medline]
50. Woods RL, Fox RM, Tattersall MH et al. Metastatic adenocarcinomas of unknown primary site. A randomized study of two combination-chemotherapy regimens. N Engl J Med 1980;303:87–89.[Medline]
51. Sporn JR, Greenberg BR. Empiric chemotherapy in patients with carcinoma of unknown primary site. Am J Med 1990;88:49–55.[CrossRef][Medline]
52. Hainsworth JD, Johnson DH, Greco FA. Cisplatin-based combination chemotherapy in the treatment of poorly differentiated carcinoma and poorly differentiated adenocarcinoma of unknown primary site: results of a 12-year experience. J Clin Oncol 1992;10:912–922.[Abstract]
53. Briasoulis E, Kalofonos H, Bafaloukos D et al. Carboplatin plus paclitaxel in unknown primary carcinoma: a phase II Hellenic Cooperative Oncology Group Study. J Clin Oncol 2000;18:3101–3107.[Abstract/Free Full Text]
54. Hainsworth JD, Erland JB, Kalman LA et al. Carcinoma of unknown primary site: treatment with 1-hour paclitaxel, carboplatin, and extended-schedule etoposide J Clin Oncol 1997;15:2385–2393.[Abstract/Free Full Text]
55. Greco FA, Burris HA 3rd, Litchy S et al. Gemcitabine, carboplatin, and paclitaxel for patients with carcinoma of unknown primary site: a Minnie Pearl Cancer Research Network study. J Clin Oncol 2002;20:1651–1656.[Abstract/Free Full Text]

This article has been cited by other articles:

				D. Talantov, J. Baden, T. Jatkoe, K. Hahn, J. Yu, Y. Rajpurohit, Y. Jiang, C. Choi, J. S. Ross, D. Atkins, et al. A Quantitative Reverse Transcriptase-Polymerase Chain Reaction Assay to Identify Metastatic Carcinoma Tissue of Origin J. Mol. Diagn., July 1, 2006; 8(3): 320 - 329. [Abstract] [Full Text] [PDF]

				J. L. Dennis, T. R. Hvidsten, E. C. Wit, J. Komorowski, A. K. Bell, I. Downie, J. Mooney, C. Verbeke, C. Bellamy, W. N. Keith, et al. Markers of Adenocarcinoma Characteristic of the Site of Origin: Development of a Diagnostic Algorithm Clin. Cancer Res., May 15, 2005; 11(10): 3766 - 3772. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) eLetters: Submit a response to this article Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services E-mail this article link to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints/Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mintzer, D. M. Articles by Greene, G. Search for Related Content PubMed PubMed Citation Articles by Mintzer, D. M. Articles by Greene, G.