Exploring Molecular Genetic Alterations and RAF Fusions in Melanoma: A Belvarafenib Expanded Access Program in Patients with RAS/RAF-Mutant Melanoma

Abstract

Background

Melanoma incidence is on the rise in East Asia, yet studies of the molecular landscape are lacking in this population. We examined patients with melanoma who underwent next-generation sequencing (NGS) at a single tertiary center in South Korea, focusing on patients harboring NRAS or RAF alterations who received belvarafenib, a pan-RAF dimer inhibitor, through the Expanded Access Program (EAP).

Patients and Methods

Data were collected from 192 patients with melanoma who underwent NGS between November 2017 and May 2023. Variant call format data were obtained and annotated. Patients in the EAP received 450 mg twice daily doses of belvarafenib.

Results

Alterations in the RAS/RTK pathway were the most prevalent, with BRAF and NRAS alteration rates of 22.4% and 17.7%, respectively. NGS enabled additional detection of fusion mutations, including 6 BRAF and 1 RAF1 fusion. Sixteen patients with NRAS or RAF alterations received belvarafenib through the EAP, and disease control was observed in 50%, with 2 patients demonstrating remarkable responses.

Conclusions

Our study highlights the value of NGS in detecting BRAF, NRAS mutations and RAF fusions, expanding possibilities for targeted therapies in malignant melanoma. Belvarafenib showed clinical benefit in patients harboring these alterations. Ongoing trials will provide further insights into the safety and efficacy of belvarafenib.

Introduction

The incidence of newly diagnosed patients with melanoma has risen each year, with 324,635 new cases and 57,043 new deaths in 2020.¹ Melanoma incidence is also on the rise in East Asia, yet a comprehensive understanding of the molecular landscape is lacking in this population.² Over the past decade, systemic therapies for malignant melanoma have evolved, with standard options including immune checkpoint inhibitors and molecularly targeted therapies for advanced-stage disease with regional or distant metastases.³ Notably, patients harboring BRAF V600 activating mutations are now endowed with expanded treatment options of BRAF and MEK kinase inhibitor combinations. Based on results from phase III trials, currently approved drug combinations include vemurafenib with cobimetinib, dabrafenib with trametinib, and encorafenib with binimetinib.^3-7 Also, given the higher prevalence of c-KIT alterations observed in the Asian populations, KIT inhibitors such as imatinib, nilotinib, and regorafenib may be considered for patients harboring c-KIT mutations.^8-12 However, for individuals with other non-BRAF mutation subtypes or NRAS mutations, targeted therapies are currently lacking, leaving most of these patients reliant on immune checkpoint inhibitors and cytotoxic chemotherapy options.

According to The Cancer Genome Atlas (TCGA) data, melanoma can be genomically subdivided into the following 4 groups based on the mutated genes; mutant BRAF, mutant RAS, mutant NF1, and triple-wild type (Triple-WT) groups.¹³ Conventionally, BRAF mutation was assessed via polymerase chain reaction (PCR)—based assay, but with the advent of expert gene panels, next-generation sequencing (NGS) has emerged as a more accessible and high-throughput method. NGS has become a cornerstone of precision medicine, prompting clinicians to use it early in cancer diagnosis and treatment on a patient-specific basis.^14,15 Compared to the conventional, single-gene targeted, PCR-based methods, NGS, which incorporates targeted RNA sequencing, also has the advantage of identifying fusion genes, particularly those with novel fusion partners not targeted by probes.¹⁶

Belvarafenib (HM95573) is a pan-RAF dimer inhibitor known to inhibit both BRAF and CRAF monomers/homodimers/heterodimers.^17,18 In phase Ib trial of belvarafenib in combination with cobimetinib for patients with advanced solid tumors harboring either NRAS Q61 or BRAF V600 mutations, promising results in terms of efficacy and tolerability were reported, including those who had previously received BRAF and MEK inhibitor combinations.¹⁹ Furthermore, recent studies have suggested that belvarafenib may have the potential for treating patients with brain metastases, as it has shown high brain/plasma concentration in preclinical models.²⁰ In line with these preclinical and clinical trial results, patients with NRAS, RAF mutations or fusions without currently available therapeutic options have been granted access to belvarafenib through the Hanmi Expanded Access Program (EAP) program in South Korea.

In this study, we conducted a comprehensive analysis of patients with malignant melanoma who had undergone NGS testing at our institution, further highlighting clinical experiences from the subset of patients who, through either PCR or NGS, were found to harbor potentially targetable mutations and received belvarafenib monotherapy through the EAP program. Here, we report real-world experiences and share perspectives on the potential of RAF fusions and NRAS mutations as clinically actionable genomic alterations.

Methods

Patient Data Collection

Data for patients with advanced malignant melanoma, diagnosed and followed up at a single academic institution (Yonsei Cancer Center, Seoul, South Korea) between November 2017 and May 2023, and whose next-generation sequencing (NGS) data were available, were obtained.

Patient demographics and clinical information were obtained from the clinical database maintained by the YCDL (Yonsei Cancer Data Library) through the medical records at the Yonsei Cancer Center, Severance Hospital. Patient characteristics including age, sex, primary tumor subtype, sites of metastases, tumor stage based on the 8th edition staging system of the American Joint Committee on Cancer (AJCC), and previous treatment including surgery, radiotherapy, and systemic treatment were collected. Results of routine molecular testing other than NGS, including BRAF and NRAS polymerase chain reaction (PCR) assays, were also noted. If available, clinical survival data including the date of disease progression after each line of therapy and the date of death were also collected.

Molecular Analysis

Next Generation Sequencing

Since November 2017, patients with advanced-stage malignant melanoma have been offered the option of NGS testing for targetable molecular alterations. Briefly, patient tumor cells were isolated from formalin-fixed and paraffin-embedded (FFPE) specimens and DNA was extracted using standard techniques. NGS was performed using a hybridization-based capture platform (TruSight Oncology 500 Kit; Illumina). Patient NGS data were obtained in Variant Call Format (VCF) files which served as the primary input data for further analysis.

Mutation Calling (Variant Annotation)

The sequenced reads were aligned to the human genome assembly (hg19) using Pisces (v.5.2.11.63). The initial aligned VCF files underwent further quality score-based filtration using SAMtools and BCF tools, applying specific criteria, including minimum read quality (read depth) and minimum variant frequency. Variants were filtered to retain those meeting stringent quality thresholds of variant quality > 20 and variant allele frequencies (VAF) > 0.1%.

To enhance the robustness of variant annotations, we used AnnoVar.²¹ gnomAD (version 2.1.1, available at https://gnomad.broadinstitute.org/),²² ClinVar (clinvar_20220320/hg19, accessible at https://www.ncbi.nlm.nih.gov/clinvar/),²³ COSMIC (cosmic v70, https://cancer.sanger.ac.uk/cosmic),²⁴ and dbSNP (dbSNP Build138, https://www.ncbi.nlm.nih.gov/snp/)²⁵ data sets were used as a reference database for known polymorphic sites and clinical significance. To ensure the exclusive inclusion of clinically relevant variants, we selectively extracted variants annotated as “Pathogenic,” “Likely pathogenic,” “Conflicting interpretations of pathogenicity,” “Drug response” from the AnnoVar files.

R-package “maftools” (version 2.14.0) was used to merge each annotated VCF file into a Mutated Annotation Format (MAF) file and for subsequent visualization and comprehensive summarization.²⁶

Pathway Map

The visualization of the enrichment pathway was carried out using pathway mapper, a web-based visualizing tool that encompasses a range of cancer-related pathways.²⁷ Each pathway data was based on TCGA Pan-Cancer pathways, and the alteration frequency data was extracted from the original MAF file.

Belvarafenib Treatment Through the EAP

Through an EAP, patients harboring RAF or RAS mutations as determined by either NGS or PCR assay, and who had experienced standard therapy failure, were granted access to single-agent belvarafenib. Patients were not required to have a measurable target lesion at the start of belvarafenib therapy.

The treatment regimen consisted of oral administration of 450 mg of belvarafenib twice daily, administered in 28-day cycles. Patients underwent regular imaging assessments, such as computed tomography (CT) or magnetic response imaging (MRI), at 8- to 12-week intervals to monitor treatment response. Response analysis was based on Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. Treatment continued until disease progression, death, or at the patient’s request to discontinue, whichever came first.

Statistical Analysis

Descriptive statistics were used to summarize patient demographics. Categorical and continuous variables were compared using the chi-square test and unpaired t tests, respectively. Progression-free survival (PFS) denoted the time from treatment initiation to disease progression or death, while overall survival (OS) measured the time from treatment initiation to all-cause mortality. Data were censored at the last observation point in case of tracking discontinuation before disease progression or death, treatment cessation due to toxicities, regimen changes, or withdrawal of consent.

Results

Baseline Characteristics

From November 2017 to May 2023, data were acquired from 192 patients with advanced-stage malignant melanoma who underwent NGS testing for targetable molecular alterations at Yonsei Cancer Center (YCC). Patient demographics and baseline clinical characteristics are described in Table 1. The median age at diagnosis was 60 years, with a slightly higher number of female (53.6%) patients. Patient with melanoma subtypes included acral melanoma in 62 (32.3%), mucosal melanoma in 61 (31.8%), and cutaneous melanoma, including both chronic sun damage (CSD) and non-CSD melanomas in 39 (20.3%) within the study population. Patients with choroidal melanoma and melanoma with unknown primary origin were also included.

At the time of molecular profiling via NGS, 76.0% (146/192) of the patients had received at least one line of systemic therapy, and 41.1% (79/192) had received 2 or more lines of therapy.

NGS Detection of Potentially Targetable Genetic Alterations: A Comparative Analysis With Conventional PCR

The spectrum of driver gene alterations, as determined by NGS, is visually represented in Fig. 1A. Special consideration was given to well-known and prevalent mutations such as BRAF and NRAS, which are potential targets for belvarafenib. We compared the mutation rates of BRAF and NRAS detected using the PCR-based and NGS assays (Fig. 1B and 1C; Supplementary Fig. S1).

Out of the 192 patients, 145 had undergone conventional, PCR-based BRAF mutation testing, with the majority being screened before NGS. The remaining 47 patients had opted for upfront NGS testing without a BRAF–PCR test. The prevalence of BRAF mutation by PCR was 16.6% (24/145). Overall, NGS detected BRAF gene aberrations in 24% of the total patients (46/192), including non-V600 mutations, fusions, and amplifications in 2 (1.0%), 6 (3.1%), and 2 (1.0%) patients, respectively. Interestingly, two patients initially classified as wild type by PCR were later revealed to have BRAF alterations by NGS, including one patient harboring BRAF amplification.

The detection of NRAS mutations using PCR was less frequent, performed in only 25% of the total patients (48/192), with a mutation incidence of 10.4% (5/48). By NGS, we detected NRAS mutations in 34 cases (17.7%), including G12/G13 point mutations, Q61 point mutations, and amplifications in 5.2%, 10.9%, and 1.6%, respectively. All NRAS mutations detected by PCR were also identified by NGS.

We noted the fraction of patient cases with at least 1 alteration in each of the following signaling pathways: RTK/RAS, cell cycle, PI3K, TP53, Notch, Wnt, Myc, Hippo, TGFbeta, and Nrf2. The most frequently altered oncogenic signaling pathway was the RTK–RAS signaling pathway (n = 100, 51.8%) which exhibited the highest median frequency of alterations among our cohort of 192 patients. Other commonly altered pathways included the TP53 pathway (8.3%), cell cycle pathway (7.3%), and PI3K pathway (7.8%). Altered pathways are depicted in Supplementary Fig. S2.

Patients With RAS and RAF Fusion—Response to Pan-RAF Inhibitor Belvarafenib

Through an expanded access program (EAP), patients with either RAF or RAS alterations were granted access to belvarafenib monotherapy after individual case evaluations. Since the first patient in January 2021, the EAP program has provided belvarafenib for 16 melanoma patients harboring various RAS and RAF mutations as shown in Table 2, including 11 patients with NRAS missense mutations, 2 patients harboring BRAF missense mutations, and 3 patients with RAF fusions (LMBR1–BRAF, AGK–BRAF, MIPOL–RAF1). All participants had undergone at least 1 prior line of palliative systemic treatment, except for 1 patient who experienced disease progression after adjuvant pembrolizumab and was subsequently treated with belvarafenib as first-line therapy. The median number of prior lines of systemic therapy was 2.5. The majority of the patients (14/16, 87.5%) had previously been exposed to immunotherapy.

At the time of data cutoff (October 26, 2023), 8 out of the 16 patients (50%) showed either partial response (PR) or stable disease (SD) and confirmed partial responses were noted in 3 patients, with 2 of these patients achieving durable responses lasting longer than 6 months (Fig. 2). Patients 2, 5, and 14, as indicated in Table 2, stopped taking belvarafenib before their response evaluation. These patients had poor performance status at the beginning of belvarafenib treatment, as determined by an Eastern Cooperative Oncology Group (ECOG) score of 2. These individuals had undergone a minimum of 3 prior therapy regimens. Apart from belvarafenib, the discontinuance occurred due to intractable acute renal failure, deterioration in mental status caused by multiple brain hemorrhagic metastases, and deteriorating concomitant renal insufficiency, respectively.

The safety profile of belvarafenib was consistent with the known safety profile of the drug. The most commonly experienced grade III adverse event (AE) was skin rash (4 cases total, 3 patients ≥ G3). Aspartate aminotransferase (AST)/alanine aminotransferase (ALT) elevation and creatinine elevation were commonly reported as well, but all reported AEs were of grades I and II.

Here, we highlight the 2 patient cases with BRAF or RAF1 fusions who had no remaining therapeutic options that have shown remarkable responses to belvarafenib.

Case 1

The first case was a 15-year-old boy with malignant melanoma with BRAF fusion mutation. The patient initially underwent left partial superficial parotidectomy due to the swelling of the parotid glands, and later through additional open biopsy, was ultimately diagnosed with primary unknown malignant melanoma spanning the left parotid gland as well as nearby neck lymph nodes. The patient received proton therapy and 3 lines of systemic therapy including pembrolizumab, dacarbazine, and paclitaxel/carboplatin until imaging follow-up revealed newly developed peritoneal carcinomatosis, liver metastasis, and increased subcarinal lymph nodes. Tissue-based NGS revealed AGK–BRAF fusion, with the preserved intact BRAF kinase domain encoded by exons 11-18 (Fig. 3A), and also FLT3 (p.T820N) and PALB2 (p.Q987*) mutations. The tumor mutation burden (TMB) was 2.4 per megabase.

Having progressed on all possible systemic treatment options and considering that he was 15 years old at the time, making him ineligible for the majority of clinical trials ongoing at that time, the patient was granted belvarafenib through the EAP program. Upon the start of belvarafenib, the patient showed a dramatic reduction in tumor size, with a notable decrease of –54.5%, and a significant decrease in the size of numerous metastatic lesions in both hemilivers (Fig. 3A). The patient was generally tolerable to the drug, experiencing only mild elevation in AST/ALT (grade I) and creatinine (grade I). The patient remained on belvarafenib and was progression-free for 16.5 months until progression in the pelvic soft-tissue lesions.

Case 2

The second case is a 67-year-old female with cutaneous melanoma initially localized in the abdomen. After undergoing an initial wide excision procedure, the disease recurred 7 months later, manifesting as multiple soft-tissue nodules in the anterior abdomen. The patient received first-line pembrolizumab with concurrent radiotherapy; however, disease progression was observed after only 3 treatment cycles. NGS analysis revealed the presence of a MIPOL1–RAF1 fusion, in concurrence with PTEN loss and CDKN2A/B loss. The kinase-coding domain of the RAF1 portion of the fusion gene remained intact (Fig. 3B).

The patient was subsequently enrolled in the EAP program and initiated on belvarafenib in April 2023. The patient experienced mild AST/ALT elevation (grade I) and skin rash (grade II), but generally adverse events were tolerable and manageable with hepatotonic agents and topical solutions. The patient demonstrated a partial response (PR), marked by a 32.1% reduction in the size of the abdomen wall lesions (Fig. 3B). The patient remains under ongoing treatment with belvarafenib, as of October 2023.

Discussion

Comprehensive genomic profiling through NGS broadens the scope of targeted therapeutics in melanoma, by identifying patients with mutations beyond the common BRAF V600E, such as gene fusions. These alterations often elude detection with traditional PCR-based assays, which, consequently, may lead to their exclusion from potential therapeutic options. We have shed light on this aspect in our retrospective analysis of the mutational landscape via NGS in melanoma, particularly focusing on the RAF and NRAS mutations.

BRAF mutations can be typically classified according to the mutation's kinase activity, RAS dependence, and dimerization status; class I mutations involve the V600 codon, class II are non-V600 mutations that are RAS-independent dimers with increased, yet somewhat weaker kinase activity, and class III mutations have low-level or impaired kinase activity, but facilitates RAS binding and CRAF activation.²⁸ These non-V600E BRAF mutations have been previously reported to be prevalent in 5% to as high as 35% of solid cancer, yet current standard targeted therapeutic options are not available for these mutations.¹⁸NRAS mutations are also prevalent in melanoma but often remain underexplored, with 1 previous phase III trial comparing binimetinib to dacarbazine in NRAS-mutant melanoma and demonstrating a modest increase in PFS for binimetinib (2.8 months vs 1.5 months).²⁹ These mutations are associated with poor prognosis yet without approved targeted therapies; MEK inhibitors have demonstrated limited clinical benefit, but resistance mechanisms involving the RTK pathway upregulation remain a challenge.³⁰

In our retrospective cohort of patients with melanoma subjected to NGS, we observed not only a strong correlation between NGS and PCR-based methods for detecting BRAF and NRAS mutations but also revealed instances of falsely identified wild-type cases by PCR, including detection of copy-number gain or amplification mutations. The prevalence of BRAF and NRAS mutation rates in our study—22.4% and 17.7%, respectively—was higher than the rates of 17.6% and 12.6% reported in a prior study involving a Korean patient with melanoma population utilizing PCR-based assays.³¹ While mutation prevalence may vary from one study to another and drawing concrete conclusions may be difficult, it is plausible to suggest that the increased prevalence of NRAS and BRAF mutation in our cohort may be attributed to the heightened sensitivity and accuracy of the NGS.

In addition to the BRAF and NRAS mutations, our NGS analysis revealed a total of 8 patients harboring either BRAF or RAF1 fusion mutations, each with distinct fusion partners (AGK–BRAF, CUX1–BRAF, TRIM24–BRAF, BRAF–DPP6, GOLG4–BRAF, LMBR1–BRAF, KIAA1548–BRAF, MIPOL1–RAF1). Notably, 1 patient exhibited concurrent BRAF–DPP6 and ESYT2–BRAF fusions.

BRAF fusions, similar to the class II BRAF mutation group, induce BRAF dimerization and constitutive activation of the MAPK pathway.^18,32 Various BRAF fusion partners have been identified in various solid cancers including melanoma, such as AGK–BRAF, KIAA1549–BRAF, AKAP9–BRAF, and TRIM24–BRAF fusions.³³ In melanoma, these fusions are more prevalent in females and can arise anywhere on the skin and mucosa.^32,34 Recent findings suggest that the emergence of BRAF fusions may play a role in resistance mechanisms across various solid cancers, including EGFR-mutant lung cancers treated with tyrosine kinase inhibitors, gastric cancer treated with FGFR inhibitors, and BRAF V600E-mutant melanomas treated with vemurafenib.^33,35,36

RAF1, also known as CRAF, belongs to the same family as BRAF and participates in the MAPK signaling pathway and is associated with various cancers including melanoma. Specific mutations, like the CRAF R391W mutation, act as driver oncogenes, promoting continuous homodimerization of the protein and increased MAPK pathway activity.³⁷ Structure variants involving the RAF1 gene fusions and their response to MEK inhibition or RAF inhibition through trametinib have been previously reported.^38,39

With the implementation of NGS to guide treatment decisions, we were able to identify patients with BRAF mutations, RAF fusions, and NRAS mutations who might potentially benefit from belvarafenib. Although our experience with belvarafenib in these patients remains limited and preliminary, we have observed encouraging results, with 50% of the patients achieving disease control. Notably, the case featuring the AGK–BRAF fusion stands out, with an exceptional tumor response to belvarafenib in a patient who might otherwise be ineligible for such clinical trials due to their young age. Also, given that the majority of patients in our study had prior exposure to immunotherapy, it appears that previous immunotherapy exposure does not act as a deterring factor in efficacy, but further research is needed to provide a more robust confirmation of this observation.

Expanding our scope to include other solid cancers, based on a case report of urothelial carcinoma featuring a NRF1–BRAF fusion that demonstrated a clinical response to trametinib,³⁹ we have enrolled a patient with renal pelvis cancer harboring NRF1–BRAF fusion into the belvarafenib EAP program. Remarkably, this patient has maintained stable disease while receiving belvarafenib for over 5 months. This observation emphasizes that response to belvarafenib extends beyond patients with melanoma, suggesting its potential efficacy in a broader range of patient populations.

Currently, there are 3 ongoing phase Ib/II trials involving belvarafenib. One phase Ib, the dose-escalation study aims to evaluate belvarafenib in combination with either cobimetinib or cetuximab in patients with locally advanced solid cancer harboring RAS or RAF mutations (NCT03284502). Another global, phase Ib trial in previous anti-PD1 or anti-PD-L1 treated, patients with NRAS-mutant advanced melanoma aims to compare belvarafenib as a single agent and in combination with either cobimetinib or cobimetinib plus nivolumab (NCT04835805). Lastly, belvarafenib in patients with BRAF class II or fusion-positive tumors is studied as part of a larger phase II, platform study (NCT04589845).

Given that the findings reported here are derived from an EAP program designed for compassionate drug use, there are some limitations. Compared with standard clinical trials, EAPs may involve fewer comprehensive data collection and reporting, and long-term follow-up data for these patients is limited, as some patients are currently still receiving treatment. However, while little is known about the long-term response and potential resistance mechanisms to belvarafenib in patients with BRAF fusions and other RAF mutations, the preliminary findings from the EAP program are promising. Furthermore, NGS in the current and upcoming decade will continue to remain a pivotal tool in identifying potential responders to targeted therapies like belvarafenib. Based on our experiences, we believe that continued clinical research with belvarafenib will significantly contribute to the advancement of precision medicine in melanoma treatment.

Conclusion

Our retrospective study has highlighted the value of NGS in detecting BRAF, NRAS mutations and RAF fusions, extending the possibilities for targeted therapies in malignant melanoma. Belvarafenib is a promising, potential treatment option for patients with these genetic alterations. Ongoing trials will provide additional insights into the efficacy and safety of belvarafenib and its role in precision medicine-based treatment strategies for patients with melanoma.

Acknowledgments

We appreciate the Medical Illustration & Design (MID) team, a member of Medical Research Support Services of Yonsei University College of Medicine, for their excellent support with medical illustrations. Belvarafenib was provided by Genentech Inc., USA, and Hanmi Pharm. Co., Ltd., Republic of Korea.

Funding

The authors declare no funding.

Conflict of Interest

Young Su Noh and Yoon-hee Hong are employees of Hanmi Pharm. Co., Ltd. The other authors indicated no financial relationships.

Author Contributions

Conception/design: K.H.K., S.J.S. Provision of study material or patients: K.Y.C., M.R.R., B.H.O., C.G.K., M.J., S.J.S. Collection and/or assembly of data: K.H.K., E.S.B., H.-J.R. Data analysis and interpretation: K.H.K., S.C., C.L., S.J.S. Manuscript writing: K.H.K., S.C., Y.J., Y.S.N., Y.-h.H. Final approval of manuscript: All authors.

Data Availability

The datasets used and/or analyzed during the study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate

This retrospective study was approved by the Institutional Review Board (IRB) at Severance Hospital, Yonsei University College of Medicine (2023-0163-002), and patient consent was waived due to the retrospective nature of the study. The study was conducted in accordance with the ethical guidelines outlined in the Declaration of Helsinki. Patient data was collected from electronic medical records and stored securely in compliance with institutional data privacy policies. All patients enrolled in the belvarafenib Expanded Access Program provided separate, written informed consent for participation in the EAP program.

Consent for publication

In accordance with ethical and legal standards, all patients enrolled in the belvarafenib Expanded Access Program provided written informed consent for the publication of their clinical data and any associated information. Patients were informed that any identifying information would be carefully anonymized to protect their privacy.

References