|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 23, No 22 (August 1), 2005: pp. 5271-5272 © 2005 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.01.4175
In Reply:University Medical Center Hamburg Eppendorf, Department of Pathology, Hamburg, Germany
King Faisal Specialist Hospital and Research Center, Biological Repository Center, Riyadh, Saudi Arabia
Institute of Pathology, University Hospital, Basel, Switzerland In a Letter to the Editor regarding our previous article1 in the Journal of Clinical Oncology, Arroyo et al describe KIT expression in three of 50 gallbladder carcinomas. Although this result is not significantly different from our previous data (zero of 20 positive; P = .15), these data raise a number of important issues. First, since immunohistochemistry (IHC) data are increasingly driving treatment decisions, it is vital to understand that IHC results from different laboratories can hardly be compared with each other. Differences in antibody selection, IHC protocols, and staining interpretation often result in highly discrepant results. This also applies for differences in tissue processing, which cannot be eliminated by using standardized US Food and Drug Administration–approved reagents. For example, Arroyo et al used a microwave slide pretreatment followed by overnight incubation. Although detailed information on critical experimental conditions (eg, incubation temperature) are not provided, the Arroyo protocol clearly differs from our procedure, which included a 3-minute pressure cooker pretreatment and 2.5 hours of room temperature incubation.1 This procedural divergence alone could explain major variations of staining results between our study1 and the study by Arroyo et al. In fact, given the notorious variability of IHC data, the major value of the data set provided in our previous study is the unusually high degree of standardization (for IHC studies), ie, the simultaneous analysis of samples from more than 120 different tumor entities in a single experiment under identical experimental conditions. This permits comparability of the expression levels and frequency between different tumor categories. In our previous study, we had described extensive experimental modifications to optimize our staining protocol.1 Although the published procedure had yielded a maximal fraction of positivity in gastrointestinal stromal tumors and minimal nonspecific background, there is no hard evidence for our protocol being optimal for selecting patients for imatinib treatment. It is therefore possible that the Arroyo protocol is "superior" to ours. The data presented in the letter by Arroyo et al strongly suggest that KIT expression can indeed occur in gallbladder cancer. This fits well with accumulating evidence suggesting that therapeutic targets that are known and established for few applications can also be present in a small fraction of many other tumor types. For example, our own recent flourescence in situ hybridization analysis of more than 300 colon carcinomas revealed few cases with classic high-level HER2 gene amplifications (unpublished data). It is therefore not unexpected that the analysis of a higher number of gallbladder cancers did identify some KIT-positive cases. As soon as targeted drugs appear on the market that are even more effective in a larger fraction of "positive cases" than for example imatinib or trastuzumab, a systematic drug target testing of all cancers, independent of their site of origin, will become an interesting option. Such a need would have dramatic logistic implications as the resources for large-scale high-quality molecular tumor tissue testing are currently nonexistent.2 Another interesting aspect of the report by Arroyo et al is the source of their tumors. The much higher incidence of gallbladder cancers in the province of Salta than in Europe3 raises the possibility of specific risk factors or genetic predispositions leading to alterations in different molecular pathways than in European gallbladder carcinomas. The different frequency of epidermal growth factor receptor exon 18-21 mutations between Japanese and Western lung cancer patients is a topical example of clinically relevant ethnic variability in drug target alterations.4 In another recent study, we had found HER2 amplification in 17% of Swiss but 31% of Saudi Arabian breast cancers.5 These differences were independent of tumor size. Importantly, such data raise the question of whether the results of studies administered on patients from the United States or Europe can always be applied to all ethnic populations. Authors' Disclosures of Potential Conflicts of Interest The authors indicated no potential conflicts of interest. REFERENCES
1. Went P, Dirnhofer S, Bundi M, et al: Prevalence of KIT expression in human tumors. J Clin Oncol 22:4514-4522, 2004
2. Sauter G, Mirlacher M: Tissue microarrays for predictive molecular pathology. J Clin Pathol 55:575-576, 2002 3. Matos EL, Loria DI, Zengarini N: Atlas de Mortalidad por Cancer en Argentina 1997-2001 [in Spanish]. Buenos Aires, Argentina, Bunge y Born Foundation, 2003
4. Paez JG, Janne PA, Lee JC, et al: EGFR mutations in lung cancer: Correlation with clinical response to gefitinib therapy. Science 304:1497-1500, 2004 5. Al-Kuraya K, Schraml P, Sheikh S, et al: Predominance of high grade pathway in breast cancer development of Middle East women. Mod Pathol 10.1038/modpathol.3800408
Related Correspondence
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2005 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
