PO.MD01.02 · 分子诊断与数据
BRAF 改变型恶性肿瘤中的动态共突变模式及治疗意义：来自 AACR Project GENIE 的证据
Dynamic co-mutation patterns and therapeutic implications in BRAF-altered malignancies: Evidence from AACR Project GENIE
作者与单位 Authors & Affiliations
摘要 Abstract
中文摘要
背景：BRAF 基因改变导致 MAPK 信号级联的持续激活，这是跨多种肿瘤类型的一个公认致癌驱动因素。在大型队列中对 BRAF 改变类型（突变、CNA 和融合）、共突变格局及可操作性进行全面的泛癌分析仍然有限。
方法：我们利用 AACR Project GENIE v18.0-public 来表征 BRAF 改变的患病率、共突变和可操作性。OncoKB 级别（L1-L4）用于定义可操作性，并对系列样本进行了克隆演化分析。
结果：在 204,292 例样本中，23,363 例检测到 BRAF 改变。12,860 例样本中鉴定出 BRAF 突变（7,359 例为 V600 变异，5,501 例为非 V600 变异）。患病率在甲状腺癌（44.9%）、黑色素瘤（38.0%）和组织细胞增生症（33.5%）中最高。V600 变异在甲状腺癌（97.2%；乳头状癌中为 98.4%，未分化癌中为 99.2%）和黑色素瘤（76.7%）中占主导，而非 V600 变异则在 NSCLC（73.3%）、子宫内膜癌（98.5%）和膀胱癌（96.6%）中富集。BRAF 拷贝数改变（CNA）见于 10,274 例样本，患病率在乳腺癌（85.9%）、食管胃癌（70.3%）和卵巢癌（69.6%）中最高。BRAF 融合发生于 1,153 例样本，患病率在胶质瘤（16.8%）和前列腺癌（14.2%）中最高。以肿瘤为中心的可操作性在组织细胞增生症（93.5% 可操作）、甲状腺癌（91.9%）和黑色素瘤（79.3%）中较高，在结直肠癌（50.6%）和胶质瘤（31.6%）中也相当可观。以改变为中心的分析证实，V600 改变在实体瘤中以及融合（例如胶质瘤中的 KIAA1549-BRAF）以 L1 证据作为最高级别。非 V600 改变则以 L2-L4 证据作为最高级别（例如组织细胞增生症中的 MEK 抑制剂；在研的 RAF 抑制剂）。非 V600 肿瘤显示出显著更高的共突变负荷，有 40 个基因富集，包括 RTK/RAS（KRAS：18.5% vs 1.3%；NF1：18.7% vs 3.6%）、染色质重塑（ARID1A：20.1% vs 7.3%）和 PI3K/AKT（PIK3CA：19.8% vs 8.8%）。共突变分析揭示了相当大比例的事件具有 OncoKB 证据，尤其是在 RTK/RAS、染色质和 PI3K/AKT 通路中，为合理的联合策略创造了机会。对来自 1,072 例患者的 2,314 例样本的系列分析表明了共突变的动态演化，支持通过系列谱分析来识别新出现的治疗靶点。
结论：本研究代表了迄今为止最广泛的 BRAF 改变泛癌分析，突出了其患病率、共突变和可操作性模式。非 V600 肿瘤表现出更高的基因组复杂性和可操作的共突变负荷，这可为联合策略提供依据。共突变的动态演化强调了通过系列谱分析来揭示新出现的治疗脆弱性的必要性。
查看英文原文 English abstract
Background: BRAF gene alterations result in persistent activation of the MAPK signaling cascade, a well-established oncogenic driver across diverse tumor types. A comprehensive pan-cancer analysis of BRAF alteration types (mutations, CNAs, and fusions), co-mutation landscapes, and actionability in large cohorts remains limited.
Methods: We utilized AACR Project GENIE v18.0-public to characterize BRAF alteration prevalence, co-mutations, and actionability. OncoKB levels (L1-L4) defined actionability, and serial samples were analyzed for clonal evolution.
Results: Among 204,292 samples, BRAF alterations were detected in 23,363. BRAF mutations were identified in 12,860 samples (7,359 with V600 and 5,501 with non-V600 variants). Prevalence was highest in thyroid cancer (44.9%), melanoma (38.0%), and histiocytosis (33.5%). V600 variants predominated in thyroid cancer (97.2%; 98.4% in papillary and 99.2% in anaplastic) and melanoma (76.7%), while non-V600 variants were enriched in NSCLC (73.3%), endometrial (98.5%), and bladder cancer (96.6%). BRAF copy number alterations (CNAs) were present in 10,274 samples, with the highest prevalence in breast (85.9%), esophagogastric (70.3%), and ovarian (69.6%) cancers. BRAF fusions occurred in 1,153 samples, with the highest prevalence in glioma (16.8%) and prostate cancer (14.2%). Tumor-centric actionability was high in histiocytosis (93.5% actionable), thyroid cancer (91.9%), and melanoma (79.3%), and was substantial in colorectal cancer (50.6%) and glioma (31.6%). Alteration-centric analysis confirmed L1 evidence as highest level for V600 alterations across solid tumors and for fusions (e.g., KIAA1549-BRAF in glioma). Non-V600 alterations showed L2-L4 evidence as highest level (e.g., MEK inhibitors in histiocytoses; investigational RAF inhibitors). Non-V600 tumors showed significantly higher co-mutation burden, with 40 genes enriched, including RTK/RAS (KRAS: 18.5% vs 1.3%; NF1: 18.7% vs 3.6%), chromatin remodeling (ARID1A: 20.1% vs 7.3%), and PI3K/AKT (PIK3CA: 19.8% vs 8.8%). Co-mutation analysis revealed a substantial fraction of events with OncoKB evidence, particularly in RTK/RAS, chromatin, and PI3K/AKT pathways, creating opportunities for rational combination strategies. Serial analysis of 2,314 samples from 1,072 patients demonstrated dynamic evolution of co-mutations, supporting serial profiling to identify emerging therapeutic targets.
Conclusions: This study represents the most extensive pan-cancer analysis of BRAF alterations, highlighting their prevalence, co-mutation, and actionability patterns. Non-V600 tumors exhibit higher genomic complexity and actionable co-mutation burden, which could inform combination strategies. The dynamic evolution of co-mutation underscores the need for serial profiling to uncover emerging therapeutic vulnerabilities.
利益披露 Disclosure
M. M. Khamis, None.