A point mutation (V600E) in the BRAF oncogene is a prognostic biomarker and may predict for nonresponse to anti-EGFR antibody therapy in patients with colorectal carcinoma. BRAFV600E mutations are frequently detected in tumors with microsatellite instability and indicate a sporadic origin. We used a mutation-specific antibody to examine mutant BRAFV600E protein expression and its concordance with BRAFV600E mutation data.
Primary stage III colon carcinomas were analyzed for BRAFV600E mutations in exon 15, and 50 BRAFV600E mutation carriers and 25 wild-type tumors were selected for analysis of BRAF proteins by immunohistochemistry (IHC). IHC was performed in archival tissue specimens using a pan-BRAF antibody and a mutation-specific antibody against BRAFV600E proteins. Staining was scored by 2 pathologists who were blinded to clinical and mutation data.
Using a pan-BRAF antibody, total BRAF protein expression was observed in the tumor cell cytoplasm in 74 of 75 colon carcinomas. A mutation-specific antibody identified diffuse cytoplasmic staining of mutant BRAFV600E proteins in 49 of 74 cancers. Analysis using a polymerase chain reaction-based assay revealed that all 49 of these cancers carried BRAFV600E mutations. In contrast, BRAFV600E staining was absent in all 25 tumors that carried wild-type copies of BRAF.
The v-Raf murine sarcoma viral oncogene homolog B1 (BRAF), a member of the RAS/RAF gene family, encodes a serine-threonine protein kinase that is a downstream effector of activated RAS. Activating BRAF mutations, characterized by a substitution of valine by glutamic acid at position 600 (V600E) on codon 15, enhance tumor cell proliferation in vitro and are detected in 8% to 15% of human colorectal cancers (CRCs).[1-3] CRCs develop through 2 major pathways that include chromosomal instability or microsatellite instability (MSI).[4, 5] BRAFV600E mutations are frequently associated with sporadic MSI tumors that arise because of epigenetic inactivation of the MLH1 mismatch repair (MMR) gene, but they are essentially lacking in tumors with germline mutations in MMR genes that cause Lynch syndrome.[5, 6] Accordingly, BRAFV600E testing has an established role in differentiating sporadic MSI CRCs from Lynch syndrome cases. Lynch syndrome, also known as hereditary nonpolyposis colorectal cancer (HNPCC), is the most common hereditary colon cancer syndrome, accounting for 3% of newly diagnosed CRC cases.
The identification of molecularly defined cancer subgroups to inform prognosis and guide targeted therapy is a goal of personalized oncology. BRAFV600E mutations are frequently detected in melanomas, and it has been demonstrated that blocking oncogenic BRAFV600E activity improves the outcome of patients with advanced melanoma.[8, 9] BRAFV600E mutations are associated with a poor prognosis in metastatic CRCs,[10-12] and most studies have reported adverse outcomes in nonmetastatic CRCs.[1, 3, 13] BRAFV600E and v-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog (KRAS) mutations are mutually exclusive in CRCs.[5, 6, 14] Evidence suggests that BRAFV600E mutations in advanced CRCs with wild-type KRAS may be associated with a lack of benefit from treatment with antibodies against the epidermal growth-factor receptor (EGFR).[2, 15] Given the prognostic and potential predictive utility of BRAFV600E status, in addition to its role in distinguishing sporadic from germline causes of MSI in CRCs, BRAFV600E has emerged as an important biomarker in clinical practice.
Currently, testing for BRAFV600E mutation status is routinely performed by DNA sequencing or using a polymerase chain reaction (PCR)-based assay. Both of these strategies are labor-intensive, relatively expensive, and depend on the variable quality of DNA extracted from formalin-fixed, paraffin-embedded (FFPE) tumor tissue. By using standard DNA sequencing, it was recently suggested that mutant BRAFV600E alleles must represent at least 25% of the signal to be reproducibly detected. These issues underscore the need for establishing a simpler and more efficient method to screen for the BRAFV600E mutation. Of particular interest is the recent development of a monoclonal antibody directed against the mutated BRAFV600E protein that can be detected by immunohistochemistry (IHC). The objective of our current study was to evaluate an anti-BRAF antibody for the detection of mutant BRAFV600E proteins and to correlate the results with mutation status by DNA sequence analysis in human colon cancers.
MATERIALS AND METHODS
Prospectively collected and pathologically confirmed TNM stage III (lymph node-positive) colonic adenocarcinomas were used for this study. All cancers were obtained from a completed phase 3 clinical trial comparing oxaliplatin, fluorouracil, and leucovorin (FOLFOX) with or without cetuximab as adjuvant chemotherapy (North Central Cancer Treatment Group Trial N0147). The study population represents a patient subset that included 50 cancers with BRAFV600E mutations and 25 tumors with wild-type (WT) copies. BRAFV600E mutation status at codon 15 was determined in extracted DNA from macrodissected FFPE tumor tissues using a multiplex allele-specific PCR–based assay and was scored for the presence or absence of the V600E variant only, as previously described. BRAFV600E testing was performed at the Mayo Clinic in a Clinical Laboratory Improvement Amendments (CLIA)-compliant laboratory. MMR protein expression (MLH1, MSH2, MSH6) had been determined in all tumors by IHC, as previously described. Tumors were classified as having deficient MMR (dMMR) if loss of 1 or more MMR proteins was detected; tumors with intact protein expression were classified as having proficient MMR (pMMR). This study and the parent clinical trial were approved by the Mayo Clinic Institutional Review Board.
IHC for BRAF protein expression was performed in FFPE tumor sections on a BenchMark XT automated slide stainer (Ventana Medical Systems, Inc., Tucson, Ariz). After deparaffinization, endogenous peroxidase activity was blocked. Optimization for each antibody included testing a range of concentrations on nonstudy cases, including mutation-negative and mutation-positive colon cancers determined by molecular testing. Staining was performed with a pan-BRAF antibody (pBR1 clone; Spring Bioscience, Inc., Pleasanton, Calif) to demonstrate total BRAF expression. Specimens were incubated with the pan-BRAF antibody (diluted 1:100) at 37°C for 16 minutes. Total BRAF protein expression was detected with the ultraView Universal 3,3 diaminobenzidine (DAB) Detection Kit (Ventana Medical Systems, Inc.), in which the ultraView Universal horseradish peroxidase (HRP) multimer was substituted for an HRP-conjugated goat anti-rat secondary antibody. We then used an anti-BRAFV600E mouse monoclonal antibody (VE1 clone; Spring Bioscience, Inc.) raised against an immunogenic synthetic peptide derived from the internal region of the BRAFV600E protein. Tissue sections were incubated with the VE1 antibody (diluted 1:45) at 37°C for 16 minutes. The immunolocalized proteins were observed using the OptiView DAB Detection Kit (Ventana Medical Systems, Inc.). After chromogenic detection, all slides were counterstained with Hematoxylin II and Bluing Reagent (Ventana Medical Systems, Inc.) for 4 minutes each, and coverslips were applied.
All immunostained slides were evaluated independently by 2 pathologists (T.C.S., S.S.) who were blinded to BRAFV600E mutation status. IHC was scored for the percentage of tumor cell immunoreactivity and staining intensity (range, 0-3+: 0, none; 1+ weak; 2+, medium; 3+ strong). For the pan-BRAF antibody, tumors was considered immunopositive when there was cytoplasmic staining of at least medium intensity (2+) above background in at least 70% of tumor cells. For the VE1 antibody, the staining characteristics are described herein. Slides were assessed for cytoplasmic staining only, and any nuclear staining or weak staining of interspersed cells was scored as negative.
Concordance between immunohistochemically analyzed BRAFV600E protein expression and BRAFV600E mutation data was analyzed. Clinicopathologic variables were investigated for associations with BRAFV600E mutational status using t tests or chi-square tests, as appropriate. P values < .05 were considered statistically significant. Analyses were performed using the SAS statistical software package (version 9.2; SAS Institute Inc., Cary, NC).
Patient Demographics and Immunostaining Results
IHC was performed to detect pan-BRAF proteins and mutation-specific BRAFV600E proteins in 75 surgically resected colon carcinomas that were previously genotyped for BRAFV600E mutation status. Patient demographics and tumor characteristics of the study population stratified by BRAFV600E mutation status are provided in Table 1. Tumor characteristics, including T classification, the number of positive lymph nodes, histologic grade, and primary tumor site, did not differ significantly according to BRAFV600E status (Table 1). Tumors with dMMR were significantly more likely to carry BRAFV600E mutations versus WT copies. The relatively high frequency of dMMR in the study population is because of enrichment with mutated versus WT BRAFV600E tumors (Table 1).
We analyzed the expression of pan-BRAF proteins in colon carcinomas to exclude the possibility that negative staining for mutant BRAFV600E proteins was falsely negative because of absent pan-BRAF expression. Using a pan-BRAF antibody, cytoplasmic staining was detected in 74 of 75 cancers in the full cohort, with 1 case removed from further testing because it was deemed nonevaluable. Among the 74 evaluable tumors, cytoplasmic staining of medium intensity (2+) for pan-BRAF proteins was detected in all tumor cells in 67 tumors (90%) and in at least 70% of tumor cells in the remaining 7 tumors (10%) (Fig. 1). Using the mutation-specific BRAFV600E antibody (VE1), cytoplasmic staining of medium or strong intensity (2+ or 3+) was detected in at least 70% of tumor cells in 49 of 74 (66%) primary colon cancers. Among these 49 tumors, 2+ or 3+ staining for BRAFV600E proteins was detected in 100% of tumor cells (n = 37), in 90% of tumor cells (n = 2), in 80% of tumor cells (n = 7), or in 70% of tumor cells (n = 3) (Fig. 1). Within these same 49 tumors, weak staining (1+) in 10% to 30% of tumor cells was observed in 11 cases (Fig. 2). Twenty-five cancers lacked any staining with the VE1 antibody but displayed diffuse staining with the pan-BRAF antibody. IHC results were interpreted independently by 2 pathologists who were blinded to BRAFV600E mutation status and to clinicopathologic variables.
Comparison of BRAFV600E Protein Expression and Genomic Mutation Testing
Our study population included 75 primary colon cancers of which 50 were BRAFV600E mutation carriers and 25 were WT tumors, as determined using a DNA-based assay. IHC results were compared with data on the mutation status of BRAFV600E in tumors. All 49 tumors that expressed pan-BRAF proteins and also had diffuse staining for mutant BRAFV600E proteins with the VE1 antibody carried BRAFV600E hotspot mutations. The 25 cancers that completely lacked BRAFV600E staining with the VE1 antibody carried WT copies of BRAFV600E.
The availability of IHC for BRAFV600E testing in archival tissue specimens can facilitate the screening of CRCs given the low frequency of BRAFV600E mutations in this tumor type. In prospectively collected primary colon cancers from a clinical trial, we observed complete concordance between mutant BRAFV600E protein expression, as detected by IHC using the VE1 antibody, with BRAFV600E mutation analysis using a PCR-based assay. Accordingly, the VE1 antibody correctly identified 49 tumors with BRAFV600E mutations and revealed a lack of staining in 25 tumors consistent with their WT BRAFV600E status. Our data are consistent with reports of BRAFV600E protein expression in primary melanomas and lung cancers, as well as in brain metastasis from multiple solid tumors, where highly concordant results with mutation testing were reported.[20-22]
In tumors with BRAFV600E mutations, immunoreactivity of medium or strong intensity using the VE1 antibody was detected in at least 70% of tumor cells in all cases and in 100% of tumor cells in 75% of these cases. Therefore, the expression of mutant BRAFV600E proteins was homogeneous in the majority of colon cancers in our study. Studies in larger numbers of tumors may provide additional information on the association of lower percentages of immunopositive tumor cells and BRAFV600E mutation status. Our findings are similar to results using the VE1 antibody in human melanomas, in which immunoreactivity was detected in greater than 80% of tumor cells in 88% of cases. Although not the situation in our study, studies in melanoma and brain metastases from multiple tumor types have observed discordant results between BRAFV600E mutation analysis and IHC using the VE1 antibody in a small number of cases.[20, 24] In melanomas, additional molecular analysis confirmed the IHC result in 3 of 5 discordant cases, suggesting that the initial molecular testing results were incorrect. To our knowledge, our data are the first to examine mutant BRAFV600E protein expression in primary CRCs. Strengths of our study include the independent scoring of BRAF protein expression by 2 pathologists who were blinded to BRAFV600E mutation status, and the fact that BRAFV600E mutation testing was performed in a CLIA-compliant laboratory.
Use of the commercially available, V600E mutation-specific VE1 antibody has the potential to streamline the process of BRAFV600E testing in clinical practice. The gold standard for detecting the BRAFV600E mutation is direct DNA sequencing. However, limitations of this approach include low tumor content, low levels of mutant DNA, and the potential for suboptimal quality of extracted DNA from FFPE tissue, although the use of microdissection and more sensitive methods have significantly improved detection rates with an ability to detect 1% to 5% of cells with the BRAFV600E mutation.[16, 25, 26] However, the assay time, expense, and the requirement for a molecular diagnostic laboratory remain limitations. In contrast, IHC offers the advantage of a relatively rapid, easy to perform, and cost-effective assay that can be readily performed by most hospital pathology laboratories.
There are at least 2 clinical scenarios in which BRAF testing can inform clinical decision-making. There is emerging consensus that all CRCs should be screened initially for MMR/MSI status in an effort to identify Lynch syndrome which has critically important implications for patients and their families. Further testing of cases with MSI and/or loss of MLH1 expression is then performed to detect a BRAFV600E mutation whose detection indicates a sporadic CRC and excludes Lynch Syndrome. As an alternative to BRAFV600E mutation testing, analysis of BRAFV600E protein expression by IHC can serve as a useful and cost-efficient screening strategy to exclude Lynch syndrome in clinical practice. In MSI CRCs found to lack mutant BRAF proteins, MLH1 methylation testing may be warranted or in the appropriate clinical setting, patients can be referred for genetic testing with DNA sequence analysis of an MMR gene(s). BRAF testing can also provide prognostic information. In the adjuvant chemotherapy trial from which our study population is derived, mutated versus WT BRAFV600E in stage III colon cancers was independently associated with significantly poorer disease-free survival. Other adjuvant studies in patients with stage II/III colon cancers and in metastatic disease have also indicated that the BRAFV600E mutation is associated with worse clinical outcome. In conclusion, our data support the use of IHC to detect mutant BRAFV600E proteins as an alternative to a DNA-based assay that can facilitate the screening of CRCs in routine clinical practice.
This work was supported in part by a National Cancer Institute Senior Scientist Award (K05 CA142885) to Dr. Sinicrope.
CONFLICT OF INTEREST DISCLOSURES
Drs. Singh, Muranyi, Shanmugam, and Grogan are employees of Roche/Ventana Medical Systems, Inc., which is the parent company of Spring Biosciences.