BRAF Mutation in Colorectal Cancer: An Update.

Colorectal cancer (CRC) is still one of the deadliest cancer-related diseases. About 10% of CRC patients are characterized by a mutation in the B-Raf proto-oncogene serine/threonine kinase (BRAF) gene resulting in a valine-to-glutamate change at the residue 600 (V600E). This mutation is also present in more than 60% of melanoma patients. BRAF inhibitors were developed and found to improve patient survival; however, most patients at the end of the track ultimately develop resistance to these inhibitors. Melanoma patients benefit from the combination of BRAF inhibitors with mitogen/extracellular signal-regulated kinase (MEK) inhibitors, among others. Unfortunately, colorectal patients do not respond much efficiently, which suggests different resistance mechanisms between the two cancer types. This review aims at shedding light on recent discoveries that improve our understanding of the BRAF mutation biology in CRC.

Introduction

Colorectal cancer (CRC) remains the third leading cause of cancer-related deaths in the United States.1 The initial theory stated by Vogelstein et al that relates the sequential multistep mutational process of CRC development is as important today as ever.2 The proposed model describes the development of colorectal tumor through sequential activation of oncogenes and concomitant inactivation of tumor suppressors.2 This model was refined recently by the so-called Big Bang model, stating that after an initial oncogenic mutation, within-tumor heterogeneity is favored by the development of subpopulations bearing other oncogenic mutational profiles.3 Mutations in the BRAF gene are examples of such oncogenic events and are found in about 10% of CRC patients.4 These mutations are associated with shorter progression-free and overall survival.5,6 BRAF inhibitors have proven high efficiency in melanoma, in which BRAF mutation rate is more than 60%, with the response rates of 50%–80% in these patients.7,8 High hopes were, therefore, raised on BRAF inhibitors for CRC BRAF-mutant treatment; however, such inhibitors revealed to be particularly ineffective with a response rate of approximately 5%.9 One of the major topical clinical challenge in therapy against CRC remains to thwart resistance to BRAF inhibitors.10 Since its discovery in 1988,11 BRAF has been more widely studied with more than 6900 related publications listed in PubMed. Nevertheless, the fact remains that the lack of knowledge about BRAF biology still precludes CRC BRAF-mutant patients from being efficiently cured. This review aims at summarizing the current state of knowledge about the BRAF mutation in CRC, with an emphasis on the very recent advances made on BRAF inhibitors in therapies as well as their resistance.

BRAF and Mutated BRAF

The serine/threonine protein kinase BRAF is an important player in the epidermal growth factor receptor (EGFR)-mediated mitogen-activated protein kinase (MAPK) pathway, where it is activated by the RAS small GTPase.12 The strength of BRAF, and also its extension to other RAF isoforms (ARAF and CRAF), is to not only activate the MAPK pathway that profoundly affects cell growth, proliferation, and differentiation but also affect other key cellular processes, such as cell migration (through RHO small GTPases), apoptosis (through the regulation of BCL-2), and survival (through the HIPPO pathway).13 Thus, it is not a surprise that BRAF is found constitutively activated by mutation in 15% of all human known cancer types.14 BRAF was reported to be mutated at several sites; however, the vast majority of mutated BRAF are V600E (1799T>A nucleotide change), characterizing up to 80% of all BRAF mutations.14 This mutation results in amino acid change that confers constitutive kinase activity.14 Most of the BRAF mutations result either in the acquirement of new phosphomimetic residues or in the release of the auto-inhibitory conformation imposed by the N-terminal region, which enhances the dimerization of the kinase domain, a crucial process for kinase activity. BRAF inhibitors have been developed by different companies; the most commonly used are vemurafenib (marketed as Zelboraf by Roche) and dabrafenib (marketed as Tafinlar by GSK), but others exist such as LGX818 (encorafenib; Novartis), XL281 (Exelixis), and CEP-32496 (Ambit Biosciences Corporation).15

Colonic polyps can be classified into adenomatous polyps (~10%) and hyperplastic polyps (~90%). Hyperplastic polyps do not progress into CRC. Some polyps are called serrated polyps (WHO classification: ICD-O 8213/0) because of their saw-toothed morphology. These polyps were long considered as nonmalignant, but this notion was challenged later. Serrated polyps are in turn subclassified into different types: serrated hyperplastic, traditional serrated adenomas (TSAs), or sessile serrated adenomas (SSAs).16 TSA and SSA are considered premalignant. The transformation of epithelia into TSA and SSA polyps has been attributed to the BRAF mutation, thus defining this mutation as an early event in progression of CRC. The activation of WNT pathway in parallel with the inactivation of p53 and p16 only appear in the late development of CRC.17 BRAF-mutated tumors are often right sided, more recurrent in woman, of higher grade, and associated with microsatellite instability (MSI) and old age.18,19 MSI is a form of genetic instability because of the deficiency of the mismatch repair machinery and results in hypermutability. MSI has been attributed to be the most prognostic factor in CRC with instability conferring a better prognosis. Interestingly, the deleterious effects conferred by the BRAF mutation were found to be more pronounced in microsatellite stable patients compared to instable ones (MSI), although not statistically different.18 The interaction between the BRAF status and the MSI status is a subject of intense debate. Proximal right-sided CRC is associated with a poorer prognosis.20 The BRAF mutation is found highly enriched in right-sided proximal tumors.21 The reason for this association is still incompletely understood.22 For more detailed descriptions on molecular mechanisms and early discoveries concerning BRAF, the reader is redirected to a recent full-depth review.12

Predictive and Prognostic Role of Mutated BRAF

While the predictive role of KRAS mutation to cetuximab (an EGFR-blocking antibody) is well established, the predictive role of mutated BRAF is a subject of intense debate. Several studies have compared that the effect of anti-EGFR was beneficial in BRAF mutants. Nevertheless, until recently, none formally studied the effect of the acquisition of the BRAF mutation on the anti-EGFR response (ie, by comparing to BRAF wild-type patients). Recently, such a study was achieved by doing a meta-analysis that grouped eight cohorts consisting of 351 BRAF-mutant patients, including BRAF wild-type patients.23 This analysis revealed that the hazard ratios of patients treated with EGFR-blocking antibodies (cetuximab or panitumumab) were not depending on the BRAF mutation status for overall survival (interaction test P-value: 0.43) but were close to significance for progression-free survival (interaction test P-value: 0.07).23 The authors concluded that the BRAF mutation was not predictive of benefits provided by anti-EGFR therapies. Similarly, another meta-analysis reported by Pietrantonio et al revealed that EGFR-blocking antibodies did not increase the efficacy of standard chemotherapy in BRAF-mutant patients.24 However, this study did not assess the survival differences between BRAF wild-type and BRAF-mutant patients.

On the contrary, the prognostic role of the BRAF mutation in CRC is well established and is usually associated with significant poorer prognosis.5,6 For example, in a study involving more than 1200 stage II and III patients, a multivariate Cox proportional hazard regression indicated that the BRAF mutation significantly affects the overall survival (hazard ratio: 1.78 [1.15–2.76]; P-value: 0.01).6 On the other hand, relapse-free survival was not found to be altered (hazard ratio: 1.30 [0.87–1.95]; P-value: 0.21).6

Resistance Against BRAF Inhibitors

The fact that the BRAF mutation is a bad predictor against anti-BRAF therapies suggests that the biology underlying this mutation is signaling through the MAPK pathway in a complex manner. Resistance to BRAF inhibitors is well known and is a subject of intense investigation. The vast majority of discoveries regarding resistance to BRAF inhibitors were achieved in melanoma because of the increased rate of BRAF V600E mutations. Currently, the field of CRC is updated with compelling discoveries.25

In melanoma, a plethora of resistance mechanisms to BRAF inhibitors were already discovered, including MEK1-, MEK2-, and NRAS-activating mutations; BRAF amplification; COT overexpression; platelet-derived growth factor receptor and EGFR overexpression; secondary RAF-related mutations; and the expression of constitutively active splicing variants of BRAF.25,26 The vast majority of these escape mechanisms still tend to be related to the overactivation of the MAPK pathway, which is why the combination of anti-BRAF and anti-MEK is an intense subject of investigation. Recently, the Food and Drug Administration (FDA) approved the use of dabrafenib (a BRAF inhibitor) and trametinib (a MEK inhibitor) for BRAF-mutated unresectable or metastatic melanomas.27

Unfortunately, discoveries made on resistance to BRAF inhibitors in melanoma may not be easily applied in CRC because of the fact that anti-BRAF drug response rates are highly different in these cancers. Also, the difference in BRAF mutation incidence in these two types of cancer may indicate that BRAF signaling is not similarly required in these two cancer types and might be context dependent. In CRC, resistance to BRAF inhibitors was shown to be driven by feedback reactivation of EGFR that activates in turn MAPK via CRAF and RAS.28,29 This activating feedback mechanism seems to be specific to CRC, as the majority of melanomas do not or only slightly express EGFR. These studies show that the feedback through EGFR could be efficiently prevented by the combination of BRAF inhibitors, such as vemurafenib and dabrafenib with EGFR and MEK inhibitors, respectively. The combination of the BRAF inhibitor with the MEK inhibitor28 or with the EGFR inhibitor29 dramatically decreases mice tumor growth or even induces tumor reduction, while the single-agent effect was only minimal. These combinations are nevertheless known to ultimately lead to resistance.30 The hepatocyte growth factor–mesenchymal–epithelial transition (MET) pathway is known to be involved in acquiring resistance to EGFR inhibitors through MET stimulation.31 Interestingly, MET inhibition can bypass this resistance.31

A very recent study unraveled unanticipated resistance mechanisms by performing exon sequencing on tumors resistant to the combination of BRAF, MEK, and EGFR inhibitors.30 This method revealed that resistant tumors exhibit KRAS and BRAF amplification as well as MEK1 mutations resulting in MAPK activation, despite the treatment with these inhibitors.30,32 Cells that became resistant to the drug cocktail maintained the BRAF mutation but displayed KRAS mutations as well (either G12D or G13D). This indicates that both mutations confer a selective advantage to the combination of BRAF, EGFR, and MEK inhibitors. However, it is well accepted that the appearance of simultaneous BRAF and KRAS mutations is an extremely infrequent event in natural colorectal tumors. The reason why it was unraveled very recently is because harboring both mutations confers negative selection for BRAF/KRAS-mutated cells that are more prone to senescence.33 In 2013, Darrin Stuart and colleagues showed that BRAF-mutant melanoma become progressively addicted to vemurafenib and that drug cessation induced tumor shrinkage.34 In this context, it could thus be appropriate to prescribe drug holidays after anti-BRAF therapy to treat BRAF-mutant patients.

Another mechanism that could explain the differential response of CRC to BRAF inhibitors versus melanoma is the levels of PI3K activation. In colorectal settings, PI3K was shown to be more consistently activated in cell lines compared to that in the melanoma settings.35 Interestingly, it was found that anti-BRAF resistance could be bypassed by PI3K inhibition. Similarly, another study showed the generation of a genetically engineered mouse model for BRAF V600E colorectal and its subsequent treatment with combinatorial anti-BRAF and dual anti-PI3K/mTOR.36 Tumor growth was found to be synergistically affected by the combination of both drugs. Altogether, the material reviewed so far suggests that the combinatorial approaches of multiple protein inhibitors in cancer therapy could be a potential strategy that raises high hopes for a better efficacy in future cancer treatments.

Future Aims and Discussion

The question of how to cure CRC BRAF-mutant patients remains unanswered at the moment but is currently in a full-scale investigation. A drug screening that was performed by Rad et al in 2013 in order to test the effects of drug combination on BRAF-mutated CRC cell lines confirmed that the combination of anti-BRAF with anti-MEK1/2 was an efficient strategy and also revealed that CHK1/2 and AURKA inhibitors, which target the cell cycle, and inhibitors of AKT, mTORC1/2, and PI3K were promising when combined with anti-BRAF.17 One major goal would be to determine why all BRAF-mutant CRC cell lines do not respond equally to the different drug cocktails. CRC BRAF-mutant patients have been considered for a long time as one subtype with high clinical similarities. However, the plethora of anti-BRAF resistance mechanisms suggest that tumors can have different overall gene expression profiles. An effort to further classify the BRAF-mutant population based on gene expression is necessary to improve personalized therapy.

Several clinical trials are currently ongoing to elucidate which treatment has the best outcome. The most awaited results come from two different ongoing clinical trials that are assessing the combination of BRAF inhibitors with MEK and EGFR inhibitors (NCT01750918 in clinicaltrials.gov) or with PI3K inhibitor (NCT01719380 in clinicaltrials.gov). A case report assessing the anti-BRAF/EGFR combination already revealed exciting results with apparent clinical effects.37 Another previous pilot study performed on 15 BRAF-mutant metastatic CRC patients with the aim to test the combination of BRAF and EGFR inhibitors revealed two things: first that this regimen was well tolerated and second that it resulted in modest efficiency, as tumor regression was observed in 10 of 12 evaluable patients.38 One should, however, keep in mind that combinatorial approaches might result in unanticipated side effects.

Altogether, recent advances in the understanding of BRAF-related biology are encouraging, and it is indisputable that we have never been closer to elucidating the resistance mechanisms of mutated BRAF to currently used anticancer drugs. This should soon result in translational facts.