Low Frequency of BRAF Mutations in Endometrial and in Cervical Carcinomas

Abstract

To the Editor:BRAF, which encodes a RAF family member that functions downstream of RAS, has been reported to be somatically mutated in a number of human cancers, most frequently at codon 599 (1). Activating mutations of BRAF have been frequently observed in colorectal carcinomas with microsatellite instability (MSI; ref. 2). In addition, mutations of BRAF and KRAS have been reported as being mutually exclusive in colorectal carcinomas (2). Mutations in BRAF have been studied in gynecological cancer. They are common in low-grade ovarian serous carcinoma, in which they provide an alternative route for activation of the RAS signaling pathway (3). In contrast, BRAF mutations have not been found in either other histological types of ovarian carcinomas or in cervical carcinomas (4).In the September 2005 issue of Clinical Cancer Research, Feng et al. reported a 21% incidence of BRAF mutations among 97 endometrial, 78 endometrioid (EEC), and 19 nonendometrioid carcinomas (5). In addition, one out of nine (11%) atypical endometrial hyperplasias also featured a BRAF mutation, most of which were located at previously unreported sites. In this series, there were no apparent differences in the prevalence of BRAF mutations among different stages, histological subtypes, and grades. The authors found 2 mutations in nonendometrioid carcinomas (11%) and 18 mutations in EECs (23%). Interestingly, BRAF mutations were more frequently found among tumors that negatively stained for hMLH1 (12 out of 32; 41%), suggesting an association between these BRAF mutations and MSI.These findings are in clear contradiction with those of three recent series exhibiting a low frequency of BRAF mutations in EEC. Thus, Mutch et al. (6) reported a single mutation among 146 EECs, which were also evaluated for KRAS mutations and MSI. This mutation occurred in 1 of the 81 MSI-positive cases. Salvesen et al. (7) studied BRAF mutations in 48 endometrial carcinomas and found one mutation in an MSI-negative EEC. In addition, Pappa et al. (4) found no mutations among the 67 endometrial carcinomas they analyzed.We have determined the incidence of BRAF mutations in exons 11 and 15 by PCR/single-strand conformational polymorphism and sequencing in a series of endometrial lesions previously characterized for KRAS, PTEN, and β-catenin mutations and MSI status (8). BRAF mutations were found in 1 of 19 atypical endometrial hyperplasias (5.6%) at codon 598 (T598I) as well as in 2 out of 94 EECs (2.1%) at codons 470 (S470F) and 614 (G614E), respectively. We also analyzed a series of 61 cervical carcinomas (19 squamous cell carcinomas and 42 adenocarcinomas) but did not find any BRAF mutations in the exons studied.With respect to cervical cancer, our results are in accordance with those reported by Pappa et al., who did not find any BRAF mutations among 47 cases, including squamous cell carcinomas and adenocarcinomas. Concerning endometrial carcinomas, our data are in close agreement with those of Pappa et al. (4), Mutch et al. (6) and Salvesen et al. (7), which indicate a low prevalence of BRAF mutation in these tumors. Differences between these series and that reported by Feng et al. (5) are difficult to explain because all of them featured an analysis of exons 11 and 15. One possible explanation may be the different ethnicity of the populations studied, whereby BRAF seems to have a more important role in Chinese patients. This is improbable, however, because the frequency of other molecular alterations, such as RAS and MSI, was very similar in all studies.In conclusion, most of the series analyzed thus far have shown a low prevalence of BRAF mutations in endometrial carcinomas, and an absence of association with other frequent molecular alterations, such as KRAS, PTEN, and MSI.In response: We thank Moreno-Bueno et al. for the letter in response to our recent article in Clinical Cancer Research (1). With regard to the differences in the incidence of BRAF mutations compared with other reports, we consider the following reasons: