Originally published as JCO Early Release 10.1200/JCO.2008.20.4297 on December 29 2008

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Hainsworth, J. D. Articles by Varadhachary, G. R. Search for Related Content PubMed Articles by Hainsworth, J. D. Articles by Varadhachary, G. R. Related Articles Related Article Social Bookmarking                            What's this?

In Reply

Sarah Cannon Research Institute, Nashville, TN

Dmitri Talantov

Veridex, Johnson & Johnson, La Jolla, CA

Martin N. Raber, Gauri R. Varadhachary

Department of Gastrointestinal Medical Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, TX

We are in full agreement with Rossi et al that morphologic examination, coupled with clinicopathologic correlation, remains the cornerstone of cancer diagnosis. In the large majority of patients, the location of the primary site is known or strongly suspected after initial clinical evaluation, and the suspected diagnosis is confirmed by morphologic examination. When the site of origin is unclear after initial clinical evaluation, the histologic appearance of the biopsy sometimes points the way to the primary site, which is subsequently confirmed by additional directed clinical evaluation. When tumors have poorly differentiated histology, immunoperoxidase staining is a valuable adjunct in identifying neoplasms with specific therapeutic implications, such as lymphoma and neuroendocrine carcinoma.

However, standard diagnostic pathologic techniques have been less successful when applied to metastatic adenocarcinomas of unknown primary site. In most cases, the light microscopic appearance is not specific enough to allow identification of the tissue of origin. Although immunohistochemical stains can often suggest a primary site, an unequivocal diagnosis is usually not possible. In fact, the immunoperoxidase studies in widespread clinical use (eg, patterns of cytokeratin [CK] 7 and CK20, and thyroid transcription factor 1) have accuracy rates of 60% to 85% when tested in cancers of known origin.^1–3 This nonspecificity is tacitly acknowledged by pathologists when they make diagnoses on the basis of immunohistochemical staining patterns. For example, a CK7+/CK20– pattern may lead to a diagnosis such as "metastatic adenocarcinoma, favor pancreas, but cannot rule out lung or breast cancer." Although second-tier immunohistochemical stains may narrow the spectrum of possibilities, the final result is not usually specific enough to provide much aid to the clinician in selecting therapy. The example given by Rossi et al emphasizes this point. Although calretinin, CK5/6, and Wilms' tumor 1 are markers most frequently associated with mesothelioma, they can also be expressed in epithelial ovarian cancer, a much more treatable cancer.^4,5

Rossi et al suggest a role for autopsies as a possible direct validation of gene profiling assays. Autopsy studies are challenging to perform, for obvious reasons, and in the era of robust and sophisticated imaging, it is hard to extrapolate data from older autopsy studies, and to reliably identify a small primary tumor in a patient with widely disseminated cancer.

Molecular profiling for unknown primary is in its infancy, and we agree that there are many unanswered questions. The molecular profiling assay is a potential new diagnostic tool, which could complement morphologic evaluation and immunohistochemistry in elucidating the tissue of origin in patients with unknown primary cancer. Recently, several such assays have been developed, which correctly identify the source of metastatic lesions in known primary cancers with a 75% to 85% accuracy rate. In the remainder of patients, results are split between wrong answers and a nondiagnostic molecular profile (ie, no diagnosis given). The reports by Horlings et al⁶ and Varadhachary et al⁷ published in Journal of Clinical Oncology discuss the first sizable experiences with molecular profiling assays in detecting the tissue of origin in patients with unknown primary cancers. (Similar studies have not been done with any of the immunohistochemical stains.) In both studies, the molecular profiling assay provided a putative tissue of origin in the majority of patients. In both reports, in most patients, the diagnoses rendered were consistent with clinical and pathologic features. These results provide indirect support for the diagnostic utility of the molecular profiling assay; direct validation of the molecular profiling assay (or any of the immunohistochemical stains) does not yet exist. In addition, as pointed out by Rossi et al, the value of molecular assays in improving the actual treatment of patients with carcinoma of unknown primary remains unproven, and there are several important questions we hope to answer in the next few years. Will a woman with an atypical metastatic pattern, in whom an ovarian site of origin is suggested by molecular profiling, have a cancer that will respond to standard ovarian cancer therapy? How should we evaluate discordant results between immunohistochemistry and a profiling assay? Is there concordance between various profiling platforms (eg, will reverse transcriptase polymerase chain reaction assays, DNA microarrays, and microRNAs produce the same results)?

As treatment for various adenocarcinomas improves and becomes more specific, accurate diagnosis in patients with unknown primary cancer will become increasingly important. Additional validation of the importance and cost effectiveness of molecular profiling assays in the diagnosis and treatment of patients with unknown primary cancer is critical. In an ongoing prospective trial, we are using the molecular profiling assay diagnosis to direct therapy for patients with unknown primary cancer, as well as studying other profiling platforms.

We believe that in the era of individualized medicine, molecular profiling will prove to have an important role in the diagnosis of patients with unknown primary cancer. However, optimal diagnosis of these challenging diseases will continue to require close communication between the clinician and pathologist.

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a "U" are those for which no compensation was received; those relationships marked with a "C" were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.

Employment or Leadership Position: Dmitri Talantov, Johnson & Johnson (C) Consultant or Advisory Role: None Stock Ownership: Dmitri Talantov, Johnson & Johnson Honoraria: None Research Funding: John D. Hainsworth, Veridex; Gauri R. Varadhachary, Veridex Expert Testimony: None Other Remuneration: None

REFERENCES

1. Tot T: Adenocarcinomas metastatic to the liver: The value of cytokeratins 20 and 7 in the search for unknown primary tumors. Cancer 85:171–177, 1999.[CrossRef][Medline]

2. Brown RW, Campagna LB, Dunn JK, et al: Immunohistochemical identification of tumor markers in metastatic adenocarcinoma: A diagnostic adjunct in the determination of primary site. Am J Clin Pathol 107:12–19, 2007.

3. Gomez-Fernandez C, Jorda M, Delgado PI, et al: Thyroid transcription factor 1: A marker for lung adenocarcinoma in body cavity fluids. Cancer 96:289–293, 2002.[CrossRef][Medline]

4. Reis-Filho JS, Simpson PT, Martins A, et al: Distribution of p63, cytokeratins 5/6 and cytokeratin 14 in 51 normal and 400 neoplastic human tissue samples using TARP-4 multi-tumor tissue microarray. Virchows Arch 443:122–132, 2003.[CrossRef][Medline]

5. Mittal K, Soslow R, McCluggage WG: Application of immunohistochemistry to gynecologic pathology. Arch Pathol Lab Med 132:402–423, 2008.[Medline]

6. Horlings HM, van Laar RK, Kerst JM, et al: Gene expression profiling to identify the histogenetic origin of metastatic adenocarcinomas of unknown primary. J Clin Oncol 26:4435–4441, 2008.[Abstract/Free Full Text]

7. Varadhachary GR, Talantov D, Raber MN, et al: Molecular profiling of carcinoma of unknown primary and correlation with clinical evaluation. J Clin Oncol 26:4442–4448, 2008.[Abstract/Free Full Text]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

Related Article

• Morphology and More Specific Immunohistochemical Stains Are Fundamental Prerequisites in Detection of Unknown Primary Cancer
  Giulio Rossi, Nazarena Nannini, and Matteo Costantini
  JCO 2009 27: 649-650 [Full Text]

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Hainsworth, J. D. Articles by Varadhachary, G. R. Search for Related Content PubMed Articles by Hainsworth, J. D. Articles by Varadhachary, G. R. Related Articles Related Article Social Bookmarking                            What's this?