BRAF-Mutated Advanced Colorectal Cancer: A Rapidly Changing Therapeutic Landscape

Abstract

BRAF-mutated advanced colorectal cancer is a relatively small but critical subset of this tumor type on the basis of prognostic and predictive implications. BRAF alterations in colorectal cancer are classified into three functional categories on the basis of signaling mechanisms, with the class I BRAF^V600E mutation occurring most frequently in colorectal cancer. Functional categorization of BRAF mutations in colorectal cancer demonstrates distinct mitogen-activated protein kinase pathway signaling. On the basis of recent clinical trials, current standard-of-care therapies for patients with BRAF^V600E-mutated metastatic colorectal cancer include first-line cytotoxic chemotherapy plus bevacizumab and subsequent therapy with the BRAF inhibitor encorafenib and antiepidermal growth factor receptor antibody cetuximab. Treatment regimens currently under exploration in BRAF^V600E-mutant metastatic colorectal cancer include combinatorial options of various pathway-targeted therapies, cytotoxic chemotherapy, and/or immune checkpoint blockade, among others. Circumvention of adaptive and acquired resistance to BRAF-targeted therapies is a significant challenge to be overcome in BRAF-mutated advanced colorectal cancer.

INTRODUCTION

Molecular profiling in metastatic colorectal cancer (mCRC) is an integral and essential component of patient care, as many genomic alterations are prognostic and/or predictive of response to therapeutic regimens. One such alteration, the BRAF gene mutation, has increased in importance in the past decade as it has strong implications for the optimization of the care of patients with mCRC. This review will discuss the biology of BRAF mutations in mCRC (particularly BRAF^V600E), recently completed and ongoing BRAF-directed clinical trials, mechanisms of resistance to BRAF-directed therapy, and future treatment directions for this important biologic subset of mCRC.

KEY POINTS

• BRAF-mutant colorectal cancer (CRC) is an important subset of CRC with distinct prognostic and therapeutic implications

• BRAF mutations in CRC are grouped into three functional classifications on the basis of underlying signaling mechanisms

• Initial treatment of BRAF^V600E-mutant metastatic colorectal cancer (mCRC) currently relies on a cytotoxic chemotherapy backbone plus bevacizumab

• Second-line recommended therapy for BRAF^V600E-mutant mCRC includes combined BRAF and epidermal growth factor receptor inhibition such as encorafenib and cetuximab per the BEACON CRC trial

• Further treatment options for BRAF^V600E-mutant mCRC are being explored, with a focus on understanding primary and acquired resistance to cytotoxic chemotherapy and targeted therapies in this disease

CONTEXT

• Key Objective

• BRAF-mutated colorectal cancer (CRC) is a complex and important subset of CRC, with BRAF^V600E mutations generally conferring a clinically poorer prognosis. Given the distinct underlying biology of BRAF^V600E-mutant colorectal cancer (CRC), it is imperative to determine mechanisms of primary and acquired resistance to standard therapies; this understanding will lead to novel treatment options.

• Knowledge Generated

• Multiple studies have determined a treatment paradigm for advanced BRAF^V600E-mutant CRC that includes initial cytotoxic chemotherapy and antiangiogenic agents, followed by combined BRAF and epidermal growth factor receptor inhibition at disease progression. Ongoing preclinical and clinical investigations are exploring further targeted therapy, immunotherapy, and chemotherapy combinations on the basis of elucidated resistance mechanisms.

• Relevance

• Establishment of signaling pathway perturbations and resistance mechanisms in BRAF-mutated CRC will lead to innovative treatment options with a goal of improving patient outcomes.

BIOLOGY OF BRAF MUTATIONS IN COLORECTAL CANCER

BRAF, a serine/threonine kinase in the RAF family, is a central node in the mitogen-activated protein kinase (MAPK) signaling pathway (Fig 1A). Its activation leads to pleiotropic effects, including cell growth, proliferation, and survival, and is therefore tightly regulated. Under physiologic conditions, growth factor receptors that sense nutrients or other growth signals at the cell surface transduce a signal to RAS to turn on this pathway that switches RAS to the activated GTP-bound state. RAS activation leads to RAF dimerization and activation and amplification of the growth signal down the MAPK pathway cascade to the nucleus.^1,2 Concomitant with pathway activation, negative regulatory signals are turned on to terminate the growth stimulus.^3,4 Thus, BRAF signaling in physiologic conditions is characterized by a need for upstream input and limited duration of activation.

FIG 1.

(A) Physiologic MAPK signaling. The MAPK cascade consists of a series of signaling steps that transduce extracellular signals to changes in gene expression to activate a program of growth, proliferation, and survival. Many steps of this pathway support amplification and specificity of the transmitted signals. The simplified schema shows that RTKs are first activated by extracellular growth factors and lead to recruitment of signaling proteins to the cell membrane and their activation, including RAS, which acts as a molecular switch that is at the on position when bound to GTP. RAS activates multiple downstream effectors, including the activation and dimerization of RAF to form RAF homo- and heterodimers. Activated RAF leads to MEK phosphorylation and activation, which leads to ERK phosphorylation and activation, and induction of a transcriptional program supporting cell growth and survival. (B) Characteristics of BRAF-mutant classes. (Left) Class I BRAF mutants, typified by BRAF^V600E, signal as constitutively activated monomers. They do not require RAS for activation. These mutants strongly activate ERK, so feedback suppresses upstream signaling; tumors with these mutants are characterized by RAS in the off GDP-bound state. (Center) Class II BRAF mutants, typified by BRAF^K601E, signal as constitutively activated mutant dimers. These also do not require RAS for activation, and feedback suppresses upstream signaling. There is likely a gradient of ERK activation and associated upstream feedback suppression across class II BRAF mutants. (Right) Class III BRAF mutants, typified by BRAF^D594G/N, amplify signaling in the pathway by binding more tightly to RAS and CRAF. They require upstream input to be oncogenic and increase ERK activation. GDP, guanosine 5' diphosphate; GTP, guanosine triphosphate; MAPK, mitogen-activated protein kinase; RTK, receptor tyrosine kinase.

Alterations in BRAF are recurrent in cancer and occur in about 12% of mCRC.⁵ A classification system has been proposed for BRAF alterations on the basis of signaling mechanisms (Fig 1B and Table 1).^6,7 To be oncogenic, alterations in BRAF must subvert the need for continuous upstream activation and overcome negative feedback signals that terminate pathway activation. Class I BRAF alterations, typified by BRAF^V600E, are the most recurrent alteration in this gene and constitute about two thirds of BRAF alterations in CRC. This alteration is strongly activating; it signals as a monomer and thus does not require upstream input and RAS activation for its activation. Furthermore, its activation is insensitive to negative feedback signals acting on upstream components of this pathway;⁸ BRAF^V600E continues to signal despite negative feedback loops that suppress receptors and RAS. All BRAF^V600 mutants identified thus far act in this way,⁹ but interestingly, the glutamic acid (E) substitution at the 600 codon has been described only in CRC and not in lysine (K) or methionine (M), which are seen in other diseases.^5,10 Class II BRAF alterations signal as RAS-independent RAF dimers and therefore also do not require upstream input and are insensitive to upstream regulation.⁹ Recurrent class II alterations in CRC include BRAF^K601E mutations and BRAF kinase fusions. Class III BRAF alterations have low activity or are kinase dead and act by amplifying an upstream signal by binding more tightly to RAS and CRAF to increase signal transduction.⁶ Thus, to be oncogenic, unlike class I or II mutants, class III BRAF alterations require a second hit upstream, which can be a KRAS or NRAS mutation or epidermal growth factor receptor (EGFR) activation (through ligand dysregulation). Recurrent class III alterations in CRC include D594, G466, and N581 mutations. Although this classification system nicely separates BRAF alterations on the basis of the functional effect, further analysis of BRAF mutants suggests that there is likely a gradient of activation across class II and III alterations. Clinical series indicate that although BRAF^V600E is nearly always mutually exclusive with other pathway alterations, class II BRAF mutants co-occur with other pathway alterations in about 10%-20% of cases.¹¹ The gradient of activation likely results from the specific mutation and how activating it is, allele-mutant dose,¹² and the particular signaling environment within that tumor.

TABLE 1.

Classification of BRAF Mutations

The presence of a BRAF^V600E or a ^non-V600BRAF alteration in CRC has prognostic and predictive significance. The spectrum of BRAF alterations and response to targeted therapies vary by tissue of origin. The colorectum, particularly the left colon, is characterized by high basal EGFR signaling,^13,14 and this presents two main challenges for targeted therapies. First, it provides a selective pressure for BRAF alterations that amplify this signal; the full range of BRAF alterations is thus seen in CRC, including a relatively high frequency of class III alterations. Second, it results in a higher threshold for pathway inhibition in CRC; that is, to elicit responses, BRAF inhibitors need to inhibit activated BRAF signaling and the upstream input from EGFR. In BRAF^V600E-mutant CRC, the upstream signaling is suppressed by negative feedback loops because of the high ERK activation.¹⁵ However, when BRAF is inhibited, ERK activation is suppressed, releasing these negative feedback loops; then, upstream receptors, predominantly EGFR, that are no longer suppressed can signal through the pathway, creating a rebound in ERK activation despite drug.¹⁶ This rebound in ERK activation is often termed adaptive resistance for its rapid time frame. Thus, inhibiting BRAF^V600E-mutant CRC requires combination therapy to target all these inputs into the MAPK pathway. In contrast to the high receptor tyrosine kinase signaling in CRC, on the other end of the spectrum is melanoma, which derives from neural crest cells, and has low receptor tyrosine kinase activity. Melanomas exhibit a high frequency of BRAF^V600E mutations and are highly sensitive to BRAF inhibitors even as single agents. As described below, combination approaches and new BRAF inhibitors have been tested and continue to be developed to overcome these challenges in targeting BRAF-mutant CRC.

BRAF^V600E MUTATIONS IN CRC: CLINICAL CHARACTERISTICS OF FUNCTIONAL CLASSES AND IMPACT ON SURVIVAL AND TREATMENT

BRAF^V600E mutations in CRC, which are associated with advanced age, right-sided colon primaries, and female gender, are also associated with decreased response to chemotherapy and poor prognosis. In first-line trials, the median survival for patients with BRAF^V600E-mutant mCRC ranges from 9 to 19 months, which is less than half as long as patients with BRAF wild-type (WT) mCRC.^17-23 In CALGB 80405, a first-line trial of infusional fluorouracil, leucovorin, and oxaliplatin (FOLFOX) or fluorouracil, leucovorin, and irinotecan (FOLFIRI) plus either bevacizumab or cetuximab, patients with BRAF^V600E-mutant tumors had a median survival of 13.5 months, compared with 30.6 months in patients with BRAF WT tumors (P = .001).²⁴ In addition, in patients with mCRC who undergo resection of a metastatic lesion, BRAF^V600E mutations are associated with increased recurrence risk.²⁵ Real-world analyses of outcomes for patients with BRAF^V600E-mutant mCRC also show a negative prognostic impact. In a single-institution study, patients with BRAF^V600E-mutant mCRC had a median overall survival (OS) of 18.2 months, compared with an OS of 41.1 months for patients with BRAF^V600E WT mCRC (hazard ratio, 2.74; P < .001).²⁶ In addition to its poor prognosis, BRAF^V600E-mutant mCRC is associated with worse health-related quality of life.²⁷ BRAF^V600E-mutant mCRC is more likely to metastasize to the peritoneum, which can lead to cancer-related pain, ascites, and high risk of gastrointestinal and genitourinary tract obstruction.²⁸

BRAF^V600E mutations are also strongly associated with deficient mismatch repair (dMMR) protein expression and microsatellite instability-high (MSI-H) disease, further highlighting the need for molecular profiling of all metastatic colorectal cancers, given that this association has prognostic and therapeutic implications. Unlike its BRAF^V600E WT counterpart, BRAF^V600E-mutated, MSI-H/dMMR CRC rarely occurs in patients with germline mutations in mismatch repair genes or Lynch syndrome. In a study of 3,063 patients, 35% of patients with dMMR mCRC had a concomitant BRAF^V600E mutation, compared with only 7% of patients with proficient mismatch repair tumors (P < .001).¹⁸ Mismatch repair deficiency can abrogate the negative prognostic impact of BRAF^V600E mutations. In the same analysis, compared with patients with BRAF^V600E-mutant and proficient mismatch repair mCRC, patients with a BRAF^V600E mutation and dMMR CRC had a better prognosis. Furthermore, BRAF mutations do not affect immune checkpoint inhibitor response. In KEYNOTE-177, a randomized phase III trial, which compared first-line pembrolizumab versus chemotherapy in previously untreated patients with dMMR/MSI-H mCRC, pembrolizumab was superior to chemotherapy even in the subgroup of patients with BRAF^V600E-mutated tumors.²⁹

Despite the strong association between BRAF^V600E mutations and poor prognosis, some patients have clinical outcomes that significantly exceed expectations. Efforts are underway to better understand the molecular drivers of prognosis and response to anti-BRAF therapies. Transcriptional analysis of BRAF^V600E-mutant CRC has revealed two molecular subtypes, BM1 and BM2. Despite worse prognosis for the BM1 subtype, BM1 colorectal cancer is more sensitive to BRAF, MEK, and EGFR inhibition than the BM2 subtype.³⁰ Further analysis reveals increased immune infiltrates in BM1 samples compared with BM2 samples, which may drive improved treatment outcomes.³⁰

As described above, BRAF mutations can be grouped into three categories on the basis of their biologic and functional characteristics. ^Non-V600BRAF mutations occur in 2% of patients with mCRC.³¹ The functional category of BRAF mutations has a significant impact on prognosis and treatment response. Compared with class I BRAF^V600E mutations, ^non-V600BRAF mutations (class II and III) are associated with younger age of onset, male gender, left-sided primaries, and better prognosis.^11,32,33 Furthermore, compared with patients with BRAF^V600E-mutant tumors, patients with ^non-V600BRAF mutations are more likely to have concomitant RAS mutations and less likely to have MSI-H/dMMR tumors.³²

Despite the poor prognosis associated with BRAF^V600E-mutant mCRC, more aggressive first-line chemotherapy does not improve outcomes. In a meta-analysis of 1,697 patients from five clinical trials comparing triplet chemotherapy (fluorouracil, leucovorin, oxaliplatin, and irinotecan [FOLFOXIRI]) plus bevacizumab with doublet chemotherapy combinations (FOLFOX or FOLFIRI) plus bevacizumab, triplet chemotherapy did not improve survival.³⁴ In addition, BRAF^V600E mutations are not associated with resistance to antivascular endothelial growth factor therapy.³⁵ On the basis of these collective findings, the first-line standard of care for BRAF^V600E-mutated mCRC should include bevacizumab in combination with a chemotherapy regimen that maximizes clinical activity and quality of life.

The anti-EGFR monoclonal antibodies panitumumab and cetuximab inhibit MAPK pathway signaling, and they are the current standard of care for the treatment of KRAS and NRAS (RAS) WT, left-sided mCRC. BRAF^V600E mutations are potent activators of the MAPK pathway, and they are associated with anti-EGFR resistance. In clinical trials comparing chemotherapy with or without anti-EGFR antibodies, anti-EGFR antibodies provided limited benefit in patients with BRAF^V600E-mutant tumors.^17,19,21,36 Furthermore, meta-analyses comparing anti-EGFR antibodies versus supportive care in patients with BRAF^V600E-mutant mCRC have found minimal single-agent activity.^37-39

Further studies have examined the impact of ^non-V600BRAF mutations on anti-EGFR antibody response. On the basis of preclinical and clinical studies, class II ^non-V600BRAF mutations are associated with anti-EGFR antibody resistance, whereas class III ^non-V600BRAF mutations remain sensitive to anti-EGFR antibodies. In a multicenter pooled analysis of 40 patients with ^non-V600BRAF mutations, the response rate to anti-EGFR therapy was 8% in patients with class II mutations and 50% in patients with class III mutations.¹¹ Class II mutations also emerge as a driver of secondary (acquired) resistance to anti-EGFR therapy.¹¹

EARLY CLINICAL TRIALS FOR BRAF^V600E-MUTANT METASTATIC CRC

The phase II FIRE 4.5 study (AIO KRK-0116), the first randomized study to prospectively study first-line therapy in BRAF^V600E-mutant mCRC, compared FOLFOXIRI plus either bevacizumab or cetuximab as first-line treatment of RAS WT, BRAF^V600E-mutant mCRC (Table 2).⁴⁰ The primary end point of objective response rate was not statistically significantly different between arms, but median PFS was improved with FOLFOXIRI/bevacizumab (10.1 months) compared with FOLFOXIRI/cetuximab (6.3 months), confirming previous meta-analysis data.

TABLE 2.

Key Completed Clinical Trials for BRAF^V600E-Mutant Metastatic Colorectal Cancer

Given the relative ineffectiveness of BRAF inhibitor vemurafenib monotherapy⁴¹ and only slightly improved effectiveness with combined BRAF and MEK inhibition with dabrafenib and trametinib⁴² in BRAF^V600E-mutant mCRC (Table 2), an observation was made as previously discussed that BRAF inhibition caused feedback upregulation of EGFR.^16,43 A phase Ib study of vemurafenib, cetuximab, and irinotecan (VIC) was then performed and demonstrated an encouraging objective response rate of 35.3% (6 of 17 patients), with a median progression-free survival (mPFS) of 7.7 months (Table 2).⁴⁴ On the basis of these findings, the randomized phase II study SWOG S1406 enrolled patients with BRAF^V600E-mutant mCRC to second- or third-line irinotecan and cetuximab with or without the BRAF inhibitor vemurafenib (Table 2).⁴⁵ The primary end point of PFS was improved with the addition of vemurafenib (mPFS 4.2 v 2.0 months, P = .001), although with a disappointing response rate of 17% versus 4% (P = .05).

The randomized phase III BEACON CRC trial enrolled 665 patients with BRAF^V600E-mutant mCRC to second- or third-line anti-BRAF/MEK/EGFR triplet (encorafenib, binimetinib, and cetuximab), anti-BRAF/EGFR doublet (encorafenib and cetuximab), or control (investigators' choice of cetuximab plus irinotecan or FOLFIRI) regimens (Table 2).⁴⁶ The objective response rate was 26.8% for the triplet arm, 19.5% for the doublet arm, and 1.8% for the control arm. The mPFS was 4.5 months, 4.3 months, and 1.5 months for triplet, doublet, and control arms, respectively. The median OS was similar for both the triplet and doublet arms at 9.3 months and 5.9 months for the control arm. Furthermore, the OS benefit for triplet and doublet arms compared with control was demonstrated in all investigated subgroups. In terms of safety, the doublet arm had numerically fewer overall toxicities, including grade 3 and higher adverse events, than either the triplet or control arm. This included a notable reduction in gastrointestinal toxicities such as diarrhea, nausea, and vomiting, as well as anemia, in the doublet arm when compared with the triplet. As a result of outcomes from this trial, the doublet encorafenib plus cetuximab regimen was determined to be a new standard of care for patients with previously treated BRAF^V600E-mutant mCRC, and this regimen received US Food and Drug Administration approval on April 8, 2020.⁴⁷

ONGOING AND FUTURE BRAF^V600E-DIRECTED CLINICAL TRIALS

Despite the impressive improvement in response rate, PFS, and survival for the combination of encorafenib and cetuximab compared with chemotherapy-based treatment in the second- and third-line settings, there remains an unmet need to increase the depth and durability of treatment response. One strategy under investigation includes first-line treatment with encorafenib and cetuximab with or without chemotherapy. In the phase II, single-arm ANCHOR CRC trial, patients with BRAF^V600E-mutant mCRC who had not received prior treatment for metastatic disease received the combination of encorafenib, binimetinib, and cetuximab (Table 2). Although the combination had a response rate of 47.8% and the study met criteria for further expansion, the median PFS was a rather disappointing 5.8 months.⁴⁸ To investigate the clinical activity of anti-BRAF treatment strategies in combination with first-line chemotherapy, the phase III, open-label, global BREAKWATER trial was launched (NCT04607421). After a safety lead-in, 870 patients with previously untreated BRAF^V600E-mutant mCRC will be randomly assigned to encorafenib plus cetuximab with or without chemotherapy or chemotherapy with or without bevacizumab, with PFS as the primary end point (Table 3).⁴⁹

TABLE 3.

Selected Ongoing Clinical Trials for BRAF-Mutant Metastatic Colorectal Cancer

On the basis of preclinical evidence of therapeutic synergy between BRAF-targeted therapies and immune checkpoint blockade,⁵⁰ several clinical trials are investigating targeted therapy and immunotherapy combinations in patients with BRAF^V600E-mutant mCRC (Table 3). A pilot trial evaluated the combination of the anti–programmed cell death protein-1 antibody spartalizumab (PDR001), dabrafenib, and trametinib in 21 patients with BRAF^V600E-mutant mCRC (4 MSI-H and 17 microsatellite stable), the majority of whom had not received prior BRAF inhibition or immunotherapy. The combination generated a response rate of 35% and a disease control rate of 75%.⁵¹ On the basis of these results, a phase II study of spartalizumab, dabrafenib, and trametinib enrolling both treatment-naïve and previously treated (including prior chemotherapy, anti-EGFR therapy, BRAF and MEK inhibition, and immunotherapy) patients with BRAF^V600E-mutant mCRC is ongoing (NCT03668431). In addition, a phase I/II trial evaluating the combination of encorafenib, cetuximab, and nivolumab in patients with microsatellite stable and BRAF^V600E-mutant mCRC who have received one or two prior lines of systemic therapy, not including BRAF, MEK, ERK, or programmed cell death protein-1 inhibitors (NCT04017650), is underway.

Studies of patients with BRAF^V600E-mutant mCRC treated with anti-BRAF therapies have identified multiple genomic alterations in the MAPK pathway (BRAF amplification, BRAF fusions, EGFR mutations, EGFR amplification, KRAS mutations, and NRAS mutations) as common drivers of acquired resistance.^52,53 Patients often exhibit more than one emergent resistance alteration, particularly when liquid biopsy is used to identify resistance.⁵⁴ These acquired mutations all reactivate MAPK signaling largely through RAF dimerization, which is not inhibited by current RAF monomer inhibitors in the clinic (dabrafenib, encorafenib, and vemurafenib). In addition, RAF monomer inhibitors have limited activity against ^non-V600BRAF mutations, which also signal as RAF dimers. Drug development efforts are underway to target RAF dimers.⁵⁵ The first strategy uses type II pan-RAF inhibitors. These inhibitors target both active RAF dimers and active RAF monomers at similar potencies, thereby preventing ERK activation. Several type II pan-RAF inhibitors are in development, including TAK-632,⁵⁶ LY3009120,⁵⁷ and AZ-628.⁵⁸ Unfortunately, clinical outcomes thus far from these novel inhibitors have been disappointing.⁵⁹ Additional inhibitors are in development, which are more selective for RAF dimers, such as Ponatinib Hybrid Inhibitor 1 (PHI1) and BGB-3245.⁵⁵ Another class of novel therapies includes paradox breakers. These inhibitors disrupt the RAF dimer interface, preventing the induction of BRAF homodimers and BRAF-CRAF heterodimers.^60,61 A phase I/II trial evaluating the safety and efficacy of the paradox breaker PLX8394 in 45 patients with BRAF mutations or fusions reported a response rate of 22%.⁶²

Since the most common mechanisms of acquired resistance to BRAF inhibition converge on reactivation of ERK, substantial efforts are underway to develop direct ERK1/2 inhibitors. The ERK1/2 inhibitor ulixertinib (BVD-523) has shown responses in patients with NRAS-mutant, BRAF^V600E-mutant, and ^non-V600BRAF-mutant solid tumors, with a manageable safety profile.⁶³ A phase II multicenter study is now underway to study ulixertinib for patients with ^non-V600BRAF-mutant solid tumors (NCT04488003; Table 3).⁶³ Additional ERK 1/2 inhibitors are in development. The ERK inhibitor GDC-0994 has also shown single-agent activity in patients with BRAF^V600E-mutant mCRC.⁶⁴

Building on the single-agent activity of ERK inhibitors, efforts are ongoing to combine ERK 1/2 inhibitors with other inhibitors of MAPK pathway signaling. HERKULES-3 is a phase Ib/II trial investigating the combination of the ERK inhibitor ERAS-007 with encorafenib and cetuximab in patients with BRAF^V600E-mutant mCRC (NCT05039177; Table 3). Another ongoing study is investigating the safety and efficacy of dabrafenib and trametinib in combination with the ERK 1/2 inhibitor LTT462 (NCT04294160). Finally, a phase II basket trial is investigating the ERK 1/2 inhibitor LY3214996 in combination with the CDK4/6 inhibitor abemaciclib (NCT02857270; Table 3).

Novel treatment strategies are also in development to target other key effectors of MAPK pathway signaling. Src homology-2 domain–containing protein tyrosine phosphatase 2 (Shp2) is a nonreceptor phosphotyrosine phosphatase that is a positive modulator or has Erk activity.⁶⁵ A preclinical study has shown that Shp2 can mediate adaptive resistance to BRAF inhibition in BRAF^V600E-mutant mCRC, especially when ERK reactivation is associated with feedback release of ERBB family or MET receptor, and that combined RAF, MEK, and Shp2 inhibition leads to sustained suppression of the MAPK pathway.⁶⁶ Given these preclinical findings, multiple Shp2 inhibitors have entered the clinic. Some of these Shp2 inhibitors are now being evaluated in patients with BRAF^V600E-mutant mCRC. A phase Ib, open-label platform study for patients with BRAF^V600E-mutant mCRC is evaluating the Shp2 inhibitor TNO155 in combination with dabrafenib and either the MEK inhibitor trametinib or the ERK inhibitor LTT462 (NCT04294160; Table 3).

Preclinical studies suggest that cotargeting MAPK and phosphatidylinositol 3-kinase (PI3K) pathway activation may lead to better activity, and BRAF and PI3K combinations have been tested in several clinical trials. In a phase Ib dose-escalation study of encorafenib and cetuximab with the PI3Kα inhibitor alpelisib conducted in patients with BRAF^V600E-mutant mCRC, the combination was tolerable, with a response rate of 18% and a PFS of 4.2 months.⁶⁷ The combination of encorafenib, alpelisib, and cetuximab was compared with encorafenib plus cetuximab in a randomized phase II trial with a response rate of 27% for the triplet regimen and 22% for the doublet of encorafenib and cetuximab.⁶⁸ In this study, the triplet regimen was associated with not only a modest increase in PFS but also increased toxicity. A separate phase I study evaluated the safety and efficacy of the BRAF inhibitor vemurafenib in combination with the PI3K inhibitor PX-866 in patients with BRAF^V600E-mutant solid tumors. Of 23 evaluable patients, seven had a confirmed partial response (objective response rate = 30%).⁶⁹ Given the limited single-agent activity of PI3K and AKT inhibitors in mCRC and the toxicity of these agents, interest in further exploration of this combination has waned.

Furthermore, the role of the PI3K pathway in mediating clinical adaptive and acquired resistance to BRAF^V600E-targeted therapy is still somewhat controversial, but the majority of evidence suggests that this pathway is not a major mediator of resistance. In the phase II trial of the BRAF inhibitor dabrafenib plus the MEK inhibitor trametinib, three of the five patients with a response to treatment had a concurrent PIK3CA mutation; the one patient with a complete response had a concurrent PIK3CA mutation; furthermore, all patients with concurrent PIK3CA mutations experienced tumor regression with treatment.⁴² On the basis of these data, the presence of concurrent PI3K pathway activation does not preclude response and BRAF/MEK inhibition, without EGFR inhibition, is sufficient to achieve regression in these tumors. In the SWOG S1406 trial comparing the triplet regimen of vemurafenib, irinotecan, and cetuximab with standard therapy with irinotecan plus cetuximab, the presence of a concurrent PIK3CA mutation was associated with a numerically improved progression-free survival (hazard ratio 0.3 for patients with concurrent PIK3CA mutation v 0.6 for patients with WT PIK3CA) although the comparison did not reach statistical significance.⁴⁵ Identified emergent alterations at resistance have converged on the ERK pathway, as described above, although one patient treated with the VIC regimen experienced an increase in the variant allelic fraction of a pre-existing PTEN mutation at progression.⁴⁴

Additional preclinical studies have identified other oncogenic pathways involved in BRAF treatment resistance. Preclinical evidence indicates that BRAF inhibition induces feedback reactivation of the Wnt/β-catenin pathway, thereby driving adaptive resistance. On the basis of these findings, a phase I study evaluated the combination of WNT974 (Wnt inhibitor), encorafenib, and cetuximab in BRAF^V600E-mutant mCRC (NCT02278133). Furthermore, R-spondin (RSPO) fusions activate the Wnt/β-catenin pathway, and these fusions are strongly associated with BRAF^V600E mutations.⁷⁰ RSPO fusions may be sensitive to Wnt pathway inhibitors,^71-73 pointing to a potential role for dual Wnt/β-catenin and BRAF inhibitor combinations.

In conclusion, BRAF-mutant mCRC is an extremely complex subtype of CRC that warrants ongoing investigation. In the past decade, significant breakthroughs have been made to understand and target this biologically distinct CRC subset, with some success. Future treatment options will be highly dependent on further elucidating the mechanisms of primary and acquired resistance to targeted therapies in this disease. With increasing knowledge of biologic underpinnings of this disease, novel translational approaches, and well-designed clinical trials, we believe that this challenge can be met.

AUTHOR CONTRIBUTIONS

Conception and design: All authors

Collection and assembly of data: All authors

Data analysis and interpretation: All authors

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

BRAF-Mutated Advanced Colorectal Cancer: A Rapidly Changing Therapeutic Landscape

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/jco/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).