Should Oncologists Be Aware in Their Clinical Practice of KRAS Molecular Analysis? TO THE EDITOR: KRAS mutations have been implicated in the pathogenesis of numerous tumors and detected in approximately 40% to 45% of colorectal adenocarcinomas. The mutational status of KRAS has emerged as a predictive factor for patients with metastatic colorectal cancer (mCRC) undergoing treatment with epidermal growth factor receptor (EGFR) –targeted therapy. In the clinical setting, a consistent correlation between presence of a KRAS mutation in codon 12 or 13 and lack of response to anti-EGFR monoclonal antibody therapy in mCRC patients has been described. 1-3 Based on these findings, anti-EGFR therapies were approved by the US Food and Drug Administration, as well as the European Me- dicinal Agency, for use exclusively in patients with mCRC harboring KRAS wild-type status and in combination with irinotecan after pro- gression or as first line with standard chemotherapy. Thus, KRAS mutational status actually represents a criterion to select patients who would benefit from anti-EGFR therapy. Point mutations in KRAS codons 12 and 13 represent two hotspots that include more than 95% of mutations in this gene. 4 Different methods for KRAS mutation analysis have been used in experimental settings or as part of clinical trials. Instead, in relation to KRAS mutational status for routine use, comparison of diverse tech- niques and their diagnostic applicability is being recently analyzed and needs further critical evaluation. Current international recommendations for KRAS mutational status detection advice well-established molecular assays use. 5 All of these tests have to be sensitive, specific, and reliable DNA-based assays. Among them, direct sequencing analysis and real-time polymer- ase chain reaction (PCR) are commonly used 5 whereas pyrosequenc- ing has recently emerged as a new powerful sequencing methodology for single nucleotide polymorphism/mutation analysis. Direct sequencing method is able to detect all the mutations in KRAS gene but it has lower sensitivity compared with other methods. Real-time PCR uses oligonucleotide primers that bind specifically to the most common mutations in codons 12 and 13 and it is sufficiently sensitive to identify KRAS mutants represented at a relatively low frequency. The detection limit of these two methods is around 20% of the mutation rate. 6 The pyrosequencing-based assay detects KRAS mutations in codons 12, 13 and it is more sensitive than traditional sequencing methods or real-time PCR being able to detect mutation rate repre- sented in fewer than 20% of the analyzed sample. Differences in detection limits and quantitation among KRAS testing platforms represent a relevant issue for pathologists to face with, since it is currently unknown what level of test sensitivity is unambiguously required to provide useful and predictive information in clinical practice. In our institution, pyrosequencing platform is routinely used to detect KRAS mutational status in patients with mCRC. We tested 100 consecutive patients and we observed a higher percentage, attested on more than 50% of KRAS mutated (codon 12/13) patients than that usually reported (40% to 45%). To explore this difference of approximately 20%, we retrospec- tively obtained archival tissue samples from patients previously re- sulted wild-type for KRAS codon 12/13 by real time-PCR. All these samples were reanalyzed by performing real time-PCR and pyrose- quencing methods at the same time, in the same laboratory and in a blinded way. We aimed to identify the diagnostic and clinical implications in using the two methods in KRAS mutational status detection as man- datory criterion to select patients to receive anti-EGFR therapy. Formalin-fixed paraffin-embedded tumor samples from 29 pa- tients with primary colorectal adenocarcinomas KRAS wild type for codon 12/13 were identified. Other main selection criteria were: tu- mor tissue availability, clinical response (partial or complete response; taking into account the best clinical response according to Response Evaluation Criteria in Solid Tumors [RECIST]) after cetuximab + irinotecan after progression on prior irinotecan-based therapy. Of 29 patients, three (10.3%) were identified as KRAS mutant in codon 12 G for 12D mutation by pyrosequencing whereas all of them were reconfirmed KRAS wild-type by real-time PCR. Consistent with the findings in the literature, mutations in codon 12 were far more common and all of them were G12D. All three patients who were mutant by pyrosequencing have been previously treated with anti-EGFR therapy according to wild-type status reported by real time-PCR. All of these patients with mutations showed partial response according to RECIST as best response during cetuximab-based chemotherapy. Moreover, the percentage of mutation rate at pyrosequencing resulted of 18.6%, 20.%, and 18.2%, respectively, which are under the general detection limit of real time-PCR. Our concerns are the clinical implications of using the most recent technologies in diagnostic rou- tine for KRAS status detection. Until pyrosequencing introduction in diagnostic setting, the detection limit of the other molecular tests was not inferior to 20%. Our experience suggests that in those patients with low KRAS mutational rate by pyrosequencing ( 20%), critical clinical interpre- tation of the result is needed. The identification of a very low frequency of KRAS mutations in patients could not influence negatively tumor anti-EGFR response, as demonstrated by our three mutated re- sponder patients. These observations would advise oncologists in considering critically pyrosequencing as routine diagnostic test to mandatory discriminate patients eligible to receive anti-EGFR therapy in clinical practice. Prospective studies using pyrosequencing are needed to evaluate the role of different percentages of mutated alleles in terms of response to anti-EGFR treatment in larger number of patients. Daniele Santini, Sara Galluzzo, Laura Gaeta, Alice Zoccoli, and Elisabetta Riva University Campus Bio-Medico, Rome, Italy JOURNAL OF CLINICAL ONCOLOGY C O R R E S P O N D E N C E © 2011 by American Society of Clinical Oncology 1 Journal of Clinical Oncology, Vol 28, 2011 http://jco.ascopubs.org/cgi/doi/10.1200/JCO.2010.32.7700 The latest version is at Published Ahead of Print on January 24, 2011 as 10.1200/JCO.2010.32.7700 Copyright 2011 by American Society of Clinical Oncology Downloaded from jco.ascopubs.org on January 4, 2012. For personal use only. No other uses without permission. Copyright © 2011 American Society of Clinical Oncology. All rights reserved.