|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2008.20.3604 on December 29 2008 © 2009 American Society of Clinical Oncology.
Morphology and More Specific Immunohistochemical Stains Are Fundamental Prerequisites in Detection of Unknown Primary Cancer
Section of Pathologic Anatomy, Azienda Ospedaliero-Universitaria Policlinico, Modena, Italy To the Editor: In Journal of Clinical Oncology, Horlings et al1 and Varadhachary et al2 reported their results on the helpful use of gene expression profiling in identification of metastatic adenocarcinomas or carcinomas of unknown primary origin from formalin-fixed and paraffin-embedded specimens. In our eyes, however, there are some points that require constructive criticism to prevent misleading messages emerging from these elegant but complex and expensive studies. The first point concerns the selected panel of immunohistochemical stains adopted to rule out the primary tumor, and to compare immunohistochemistry (IHC) and molecular profiling results. Horlings et al adopted a panel of 10 immunostains, but the degree of specificity of some of the employed markers—such as carcinoembryonic antigen (positive in almost all types of adenocarcinomas, and practically yielding nonspecific results) and clustered antibody 10 (positive in renal cell carcinoma, but also in several other epithelial, mesenchimal, and lymphoid neoplasms)—is poor, whereas other more specific markers were not included, such as caudal-type homeobox transcription factor 2 (a specific transcription factor of intestinal differentiation in GI tumors or adenocarcinomas from other sites with intestinal differentiation), Wilms' tumor 1 (useful in differential diagnosis between adenocarcinoma and mesothelioma, and relatively specific for gynecologic adenocarcinomas), and calretinin (positive in mesothelioma and adrenal carcinoma, but not in adenocarcinomas that often occur with mesothelioma in differential diagnosis). The choice of the IHC panel may have significantly influenced comparisons of IHC and gene expression profiling results. Another unclear point in these reports concerns who or what the discriminating factor of tumor origin was in the end. Was gene expression signature the boundary of comparison? Additional retrospective studies clearly evidencing primary tumors using autopsy examination could be of some interest to better elucidate the role of gene signatures in this setting. For instance, we found bizarre the case of patient 106 listed in Table 6 of the report by Horlings et al, in whom the gene expression profile was consistent with rhabdomyosarcoma (in a man 74 years of age), and immunohistochemical stains highlighted positivity for cytokeratin (CK) 7 and CK20 as well as carcinoembryonic antigen, among others. Again, as listed in Table 4 of the report by Vardhachary et al, a diagnosis of poorly differentiated carcinomas was made in two tumors expressing more than two markers of relatively specific mesothelial differentiation (calretinin plus CK5/6, and Wilms' tumor 1 plus calretinin), making a diagnosis of mesothelioma consistent in these patients. In addition, even when IHC suggested two or three primary tumor possibilities, major discrepancies in therapeutic strategies were limited to a couple of patients. Finally, in our experience, we have had difficulty finding patients in whom a close clinicopathologic correlation coupled with adequate immunohistochemical stains and imaging studies failed to demonstrate primary tumor, or at least to permit two or three diagnostic suggestions without dramatic changes in therapeutic options (ie, advanced cancer from upper GI tract, pancreas, stomach or biliary tract, and gallbladder). In the end, although we understand the value of gene expression profiling as one of the novel methods in understanding the mechanisms underlying diseases, in tailoring management of patients with neoplastic or non-neoplastic diseases, and in predictive and prognostic fields, as pathologists, we are also worried about these molecular studies, because they cannot take into consideration the most imponderable and key diagnostic factors—namely, morphologic features. Provided that we place morphology and molecular studies side by side even more closely in the future, we advise clinicians and young pathologists that a careful morphologic examination of well-fixed and well-performed hematoxylin-eosin–stained slides by a prepared and experienced pathologist is indispensable and invaluable, and sometimes also enables the anticipation of new tumor entities characterized by peculiar molecular characteristics.3 AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST The author(s) indicated no potential conflicts of interest. REFERENCES
1. 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. 2. 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. 3. Rosai J: Why microscope will remain a cornerstone of surgical pathology. Lab Invest 87:403–408, 2007.[CrossRef][Medline]
Related Articles
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2009 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
