Emerging Treatment Options for BRAF-mutant Colorectal Cancer

UNIQUE CLINICAL PRESENTATIONS

In the early days of understanding BRAF mutant CRC, the clinically defining feature of BRAF^V600Emutation in CRC was its dismal prognosis. We first observed these patients to be at risk for clinical deterioration after rapid sequential progression through multiple lines of standard chemotherapy. Many patients died while in queue for the first clinical trials offering a hope of personalized therapies.

Since then, we have identified multiple other clinicopathologic features often associated with the BRAF^V600E mutation. Older age, female gender, and White race have been associated with increased likelihood of harboring a BRAF^V600E mutation.^16–18 Over two-thirds of BRAF^V600E tumors originate in the right colon vs the left colon (68 vs 32%); this ratio is in contrast to the typical distribution for BRAF-wild-type tumors, which occur more frequently in the left colon (35 vs 65%; P<0.001).¹⁹ An analysis of sites of metachronous metastases found that BRAF^V600E tumors are associated with more frequent peritoneal metastases (26 vs 14%; P<0.01) and less frequent liver-limited metastatic disease (41 vs 63%; P<0.01) at the time of diagnosis.²⁰ When observing the entire disease course, significantly higher rates of peritoneal metastases (46 vs 24%; P=0.001) and distant lymph node metastases (53 vs 38%; P=0.044) and lower rates of lung metastases (35 vs 49%; P=0.049) have been reported in BRAF^V600E tumors.¹⁹

A retrospective study that matched patients with BRAF^V600E-mutant tumors to patients with BRAF-wild-type tumors based on tumor location found more peritoneal metastases and ascites in the BRAF^V600E-mutant tumors. However, when limiting this comparison to only patients with right-sided primaries, there was no difference in sites of metastasis.²¹ Notably, the increased incidence of peritoneal metastases and ascites represents unique challenges in the interpretation of responses using response evaluation criteria in solid tumors (RECIST 1.1) in the context of clinical trial evaluations.²¹

THE PROGNOSTIC AND PREDICTIVE VALUES OF A BRAF^V600EMUTATION

Following early clinical observations of the unique disease biology and dramatically inferior survival of patients with metastatic BRAF^V600E-mutant CRC, several subset analyses from multiple large clinical trials and retrospective series sought to further characterize features of BRAF^V600E-mutant CRC. In the MRC COIN trial of a fluoropyrimidine and oxaliplatin with or without cetuximab, the median OS across treatment arms was 20.1 months for patients with RAS/BRAF-wild-type tumors vs 8.8 months for patients with tumors harboring BRAF^V600E mutations.²² Among patients treated with infusional 5-fluorouracil (5-FU), leucovorin and irinotecan (FOLFIRI), and bevacizumab in the FIRE-3 study, BRAF^V600E-mutant tumors were found to have a significantly shorter OS vs those with BRAF-wild-type tumors (13.7 vs 20.1 months).²³ Cancer and Leukemia Group B (CALGB) 80405 was a recent study of the chemotherapy backbones infusional 5-FU, leucovorin and oxaliplatin (FOLFOX), or FOLFIRI with the addition of either bevacizumab or cetuximab in metastatic CRC.²⁴ Among patients receiving chemotherapy plus bevacizumab, a subset analysis demonstrated a median OS of 17.4 months in patients with BRAF^V600E mutations vs 35.1 months in BRAF-wild-type patients.²⁵

Outside of the clinical trial setting, a retrospective study of patients at two major academic centers showed a decreased median OS in patients with BRAF^V600E mutations of 10.4 vs 34.7 months (P<0.001).¹⁹ Additionally, an analysis of patients with metastatic CRC in the Nurses’ Health Study and Health Professionals’ Follow-up Study found that among patients with microsatellite stable (MSS) tumors, a BRAF^V600E mutation was associated with a colon cancer-specific mortality hazard ratio (HR) of 1.60 (95% CI =1.12–2.28; P=0.009).²⁶

Because BRAF testing is not routinely performed in nonmetastatic CRC, less is known about the prognostic and predictive values of BRAF mutations in this setting; however, retrospective studies from adjuvant chemotherapy trials have provided some insight. In a retrospective analysis of tumor samples collected from stages II and III CRC patients in the adjuvant PETACC-3 study, BRAF^V600E mutations were identified in 7.9% of patients,²⁷ similar to what is seen in metastatic disease. BRAF^V600E mutation was associated with decreased OS (HR 1.66; 95% CI 1.15–2.40) in both stages II and III patients, with an even more pronounced effect in the MSS cohort with BRAF^V600E mutations (HR 2.19; 95% CI 1.43–3.37). A similar analysis of tumor samples from 2720 stage III patients treated on the NCCTG N0147 adjuvant study found that BRAF^V600E mutation was associated with worse 5-year disease-free survival (HR 1.43; 95% CI 1.11–1.85).²⁷ In a retrospective study of 364 stages II and III patients treated in the Netherlands, a BRAF^V600E mutation was detected in 22% of stage II and 19% of stage III patients.²⁷ In a multivariate model, BRAF^V600E was similarly associated with worse OS (HR 0.45; 95% CI 0.25–0.8) and cancer-specific survival (HR 0.47, 95% CI 0.22–0.99). Based on these studies, it is thought that BRAF^V600E mutation remains associated with poor prognosis in stages II and III CRC; however, we do not have data to substantiate whether or not it is predictive of benefit from adjuvant chemotherapy.

Because RAS and BRAF^V600E mutations are typically mutually exclusive at diagnosis, patients with BRAF^V600E mutations were naively included in clinical trials of the anti-epidermal growth factor receptor (EGFR) therapies, cetuximab and panitumumab, which recruited patients with KRAS-wild-type tumors. An initial pooled analysis of 70 cases with BRAF^V600E mutations from the CRYSTAL and OPUS randomized clinical trials found that the addition of cetuximab to first-line chemotherapy was possibly associated with a nonsignificant trend toward improved progression-free survival (PFS) and OS for BRAF^V600E-mutant metastatic CRC, although inferior OS was observed across treatment arms in comparison to BRAF-wild-type patients.²⁸

However, a large meta-analysis of 463 BRAF^V600E tumors from one Phase II and nine Phase III trials of cetuximab or panitumumab did not show a PFS (HR 0.88; 95% CI 0.67–1.14; P=0.33) or an OS (HR, 0.91; 95% CI 0.62–1.34; P=0.63) benefit compared to the control chemotherapy arms.²⁹ Another meta-analysis of 351 BRAF^V600E mutant tumors from eight randomized trials (seven with OS data) similarly did not identify an advantage for anti-EGFR therapy over the control arm for either PFS (HR 0.86; 95% CI 0.61–1.21) or OS (HR 0.97; 95% CI 0.67–1.41).³⁰ Although the interaction between BRAF^V600E status and treatment response was not significant (P=0.43) and the authors concluded that the results could be due to chance alone,³⁰ the preponderance of data suggests that patients with BRAF^V600E mutations are unlikely to derive significant benefit from the currently approved anti-EGFR regimens. In line with this, current National Comprehensive Cancer Network (NCCN) consensus-based guidelines advise that the presence of a BRAF^V600 mutation confers a very low likelihood of response to anti-EGFR therapy.³¹

REVISITING A ROLE FOR CYTOTOXIC CHEMOTHERAPY

Secondary analyses of BRAF^V600E-mutant subsets from multiple randomized clinical trials demonstrating a lack of therapeutic benefit and bleak prognosis with conventional cytotoxic chemotherapy doublets³² subsequently spurred attempts at identifying alternative treatment options for patients with metastatic BRAF^V600E mutant CRC. Given the aggressive nature of BRAF^V600E disease, clinicians are often faced with the challenge of achieving initial disease control in patients at risk for rapid clinical deterioration.³³

The TRIBE study evaluated FOLFIRI plus bevacizumab vs infusional 5-FU, leucovorin, and oxaliplatin and irinotecan (FOLFOXIRI) plus bevacizumab.³⁴ Of the 508 patients in this study, 28 patients with BRAF^V600E mutations were enrolled, of whom 12 patients were assigned to the FOLFIRI arm and 16 patients were assigned to the FOLFOXIRI arm. Across both arms, the median OS in the RAS– and BRAF-wild-type patients was 37.1 vs 13.4 months in the small subset of patients with tumors harboring BRAF^V600E mutations (HR 2.79; 95% CI 1.75–4.46; P<0.0001).³⁵ Although the number of patients with BRAF^V600E mutations in this study was small, the median OS of patients treated with FOLFOXIRI plus bevacizumab in TRIBE was 19.0 months compared to 10.7 months in the FOLFIRI plus bevacizumab arm (HR 0.54; 95% CI 0.24–1.20).³⁵ An overall response was reported in 56% of patients with a BRAF^V600E mutation receiving FOLFOXIRI plus bevacizumab vs 42% of patients receiving FOLFIRI plus bevacizumab (odds ratio [OR] 1.82, 95% CI 0.91–2.62). There was no significant association between treatment effect and BRAF mutation status either for OS (P-value =0.52) or for PFS (P-value =0.68),³⁵ suggesting that patients with BRAF^V600E mutations derived equivalent benefit from FOLFOXIRI over FOLFIRI as their BRAF-wild-type counterparts. In the CHARTA Phase II study, in patients with BRAF^V600E mutations, the median PFS was 10.1 months among patients who received FOLFOXIRI plus bevacizumab vs 7.8 months in patients who received FOLFIRI plus bevacizumab (HR 0.72, P=0.61).³⁶ Acknowledging the limitations of the small sample size in both of these studies and the lack of statistical significance of the trend toward benefit, FOLFOXIRI plus bevacizumab is an option for patients with known metastatic BRAF^V600E CRC with well-preserved performance status for whom early disease control is urgent and imperative. Additionally, based on data from Loupakis et al¹⁷ showing that female patients with right-sided primary tumors with mucinous histology have an 81% chance of having a BRAF^V600E mutation, we consider upfront triplet therapy in patients who meet some of these criteria, while molecular analyses are pending.

Although BRAF mutation status was not assessed at the time of the clinical trials that led to the approval of bevacizumab in metastatic CRC, it is generally thought that BRAF^V600E does not preclude response to bevacizumab. As noted earlier in the TRIBE study, significant response has been seen with triplet chemotherapy in combination with bevacizumab.³⁴ Additionally, a subgroup analysis of BRAF^V600E-mutated tumors from CAIRO3 showed a preserved benefit from bevacizumab containing maintenance therapy compared to no maintenance as compared to patients with wild-type tumors.³⁷However, bevacizumab is highly unlikely to overcome the negative prognostic effect of a BRAF^V600Emutation; a retrospective study of patients treated with FOLFOX plus bevacizumab found the overall response rate in patients with a BRAF^V600E mutation to be only 18.4%.³⁸

DEVELOPING A PERSONALIZED APPROACH

Since BRAF^V600E CRC emerged as a distinct entity, multiple clinical trials have sought to replicate the successes of targeted therapies seen in melanoma (Table 1). The two BRAF inhibitors, vemurafenib and dabrafenib, revolutionized the treatment of BRAF^V600E metastatic melanoma and have been shown to be active in BRAF^V600E-mutated non-small-cell lung cancer, anaplastic thyroid cancer, cholangiocarcinoma, Erdheim–Chester disease, and Langerhans cell histiocytosis.³⁹ Additionally, single-agent vemurafenib was shown to arrest cell proliferation and inhibit tumor growth in CRC cell lines and xenograft models expressing BRAFV600E, respectively.⁴⁰ Based on this, a Phase II pilot study of vemurafenib was completed in patients with BRAF^V600E metastatic CR using the maximum tolerated dose of 960 mg twice daily identified in melanoma studies.⁸ Of the 21 patients enrolled, a partial response was documented in one patient and seven patients were reported to have stable disease; however, the median PFS of 2.1 months did not suggest durable activity from single-agent BRAF inhibition. Similar to the rates seen in melanoma treatment, 15 of the 21 (71%) patients experienced a grade 3 adverse event, including development of five squamous cell carcinomas of the skin, a known adverse event associated with single-agent BRAF inhibition.

(To view a larger version of Table 1, click here.)