Activating BRAF mutations are observed in malignant melanomas, pleomorphic xanthoastrocytomas, hairy cell leukemia, papillary thyroid carcinomas, colorectal cancer, ovarian cancer and ganglioglioma [1, 10, 11]. Generation of a monoclonal antibody specific for the BRAF V600E mutation (VE1) was published recently [3]. This antibody has been successfully validated for melanomas [5], lung adenocarcinomas [8] and thyroid carcinomas [2]. Although pituitary adenomas account for up to 10 % of all intracranial tumors [4], there is still a lack of data on VE1 immunostains in these tumors. In three studies in which BRAF in pituitary adenomas was examined by sequencing, a single V600E mutation was found in 145 cases [6, 7, 10].

We screened 78 pituitary adenomas (epidemiological data in supplemental Table 1) with VE1 antibody. Staining was observed in 18 cases (12 weak (15 %); 5 moderate (6 %) and 1 strong case (1 %); supplemental Figure S1). Twelve of the positive cases were HGH-producing adenomas, two were mixed mammosomatotroph adenomas and the four remaining adenomas expressed ACTH (including the case with score 3). To rule out non-specific enzymatic activity of the avidin–biotin method, the VE1 diaminobenzidine (DAB) staining results were repeated using two different polymer detection systems (Ventana OptiView and Leica BondMax), as recommended by Andrulis et al. [1] to reduce unspecific background staining, and reproducibly confirmed the positive reactions (Fig. 1b). The signal was even stronger than in the initial DAB staining of the same case (Fig. 1a). The cases were also stained with synaptophysin and CD68; possible occasional cross-reactivity was excluded [3]. Ten VE1 positive cases (one score 3, three score 2 and six score 1) were analyzed by clamped probe assay; they showed wild-type BRAF curves (Fig. 1c; for experimental methods see supplemental data). Sanger sequencing of BRAF exon 15 in the same cases confirmed the absence of V600E mutation (Fig. 1d). We minimized sequencing failure because we used only samples with 80 % tumor content and additionally performed clamped probe assays that are able to detect mutations even if the mutation/wild-type ratio is as low as 1 % [9]. Due to their high-sequence homology with the BRAF activation segment, we also analyzed the activation segments of ARAF and CRAF by sequencing to rule out VE1 cross-reactivity with a putative homolog ARAF V453E or CRAF V492E mutation, again revealing wild-type sequences for all cases (Fig. 1d; for experimental procedures, see supplemental data).

Fig. 1
figure 1

Strong unspecific immunohistochemical staining [avidin–biotin, DAB (a) and polymer-based Ventana OptiView (b)] of VE1 in a BRAF wild-type ACTH-producing adenoma. BRAF clamped probe assay shows the wild-type melting curve for this case. 1 pmol locked nucleic acid (LNA) suppresses the wild-type BRAF to negative control levels (c). Sequencing of ARAF, BRAF and CRAF activation segments reveals wild-type DNA sequence. Black rectangles mark the phosphorylation sites of the activation segments. The BRAF V600 Codon and respective V453 and V492 homologs of ARAF and CRAF are marked with red rectangles (d)

Full size image

In contrast to many previously investigated cancers, our results indicate that even strong positive immunostaining of VE1 in pituitary adenomas does not indicate the presence of a BRAF V600E mutation, especially if it is restricted to the cells’ membranes. While V600E mutations are very rare, there is evidence of overexpression of BRAF in non-functioning pituitary adenomas [7]. As the VE1 immunoreactivity in pituitary adenomas was restricted to tumor cells and did not correlate to overall pan-BRAF pBR1 immunostaining (supplemental data), this non-specific VE1 immunoreactivity runs the risk of false positive interpretation.

Our data demonstrate that, while VE1 may represent a specific tool for detecting BRAF V600E mutations in other tumors, it is unsuitable for detecting BRAF V600E in pituitary adenoma. Thus, the specificity of VE1 should be thoroughly examined for each tumor entity by paralleled genetic mutation analysis prior to routine application for research or diagnostics.