PO.CL06.02 · 临床研究
超突变模式塑造错配修复缺陷型胶质瘤的肿瘤发生与免疫治疗反应
Hypermutation patterns shape tumorigenesis and immunotherapy response in mismatch repair deficient glioma
作者与单位 Authors & Affiliations
摘要 Abstract
中文摘要
原发性错配修复缺陷型(priMMRD)胶质瘤是致命的癌症,占儿童和青年胶质瘤的5-10%。priMMRD胶质瘤的特征是普遍的超突变、对化疗-放疗的耐药,以及对抗PD1免疫治疗显著但异质性的反应。为阐明复制错误驱动的诱变对priMMRD胶质瘤演化及其异质性临床病程的影响,我们分析了来自国际复制修复缺陷联盟(International Replication Repair Deficiency Consortium)的大型priMMRD胶质瘤队列(n=162)的基因组、甲基化组、免疫及临床数据。相较于非MMRD胶质瘤,MMRD导致高微卫星不稳定性和超突变,从而引起全局性DNA低甲基化,并可能影响其分类。复制型聚合酶和IDH1中反复出现的体细胞驱动突变将priMMRD胶质瘤分层为3个独特的遗传学和临床亚组:priMMRD1(MMRD+ POLE/POLD1,56%)、priMMRD2(仅MMRD,27%)和priMMRD3(MMRD+ IDH1,17%)。priMMRD1胶质瘤具有完全的复制修复缺陷,起源于胚系双等位基因MMRD(CMMRD),发生于低龄,并表现出超高突变。priMMRD2胶质瘤富集于单等位基因胚系MMRD(Lynch综合征)和较高龄。最后,IDH1驱动的priMMRD3胶质瘤携带MutSalpha(MSH2/MSH6)的胚系突变、较低的突变负荷及独特的影像学模式。各亚组还利用不同的基因组不稳定性机制:priMMRD1胶质瘤在胶质瘤驱动基因中携带频繁的点突变,而priMMRD2和priMMRD3胶质瘤在相同基因中具有较高的拷贝数改变率,包括CDKN2A(P<0.0001)和PTEN(P<0.0001)。在MMRD驱动的胶质瘤发生中观察到的驱动突变,可用构成MMRD突变特征的特定三核苷酸背景来解释。TP53和IDH1中的热点突变发生于MMRD常突变的背景中,而不源自频繁突变背景的儿童胶质瘤驱动基因(BRAF/H3-3A)则罕见。利用突变特征和变异等位基因分数,我们构建了一个MMRD胶质瘤发生模型,其中TP53突变最先发生,随后是POLE/POLD1、IDH1及其他胶质瘤驱动基因的继发突变。priMMRD胶质瘤亚组在免疫微环境和免疫治疗反应方面存在差异:priMMRD1胶质瘤与抗PD1单药治疗下最高的2年总生存率(75%)相关,表现出促免疫表达特征和高CD8+ T细胞浸润,与priMMRD2(24%)和priMMRD3胶质瘤(0%)形成对比。这些发现提示,MMRD诱变通过驱动突变的获得及免疫治疗反应塑造了priMMRD胶质瘤的独特图谱,为亚组特异性免疫治疗组合的优化提供了合理的方法。
查看英文原文 English abstract
Primary mismatch repair deficient (priMMRD) glioma are deadly cancers comprising 5-10% of gliomas in children and young adults. PriMMRD glioma are characterized by universal hypermutation, resistance to chemo-irradation, and striking, but hetergeonous, responses to anti-PD1 immunotherapy. To elucidate the impact of replication-error driven mutagenesis on priMMRD glioma evolution and their heterogenous clinical courses, we analyzed genomic, methylomic, immune, and clinical data on a large cohort of priMMRD glioma (n=162) from the the International Replication Repair Deficiency Consortium. MMRD contributed to high microsatellite instability and hypermutation resulting in global DNA hypomethylation compared to non-MMRD gliomas and potentially affecting their classification. Recurrent somatic driver mutations in replicative polymerases and IDH1 stratified priMMRD gliomas into 3 distinct genetic and clinical subgroups: priMMRD1 (MMRD+ POLE/POLD1 , 56%), priMMRD2 (MMRD-only, 27%), and priMMRD3 (MMRD+ IDH1 , 17%). PriMMRD1 gliomas, which harbor complete replication repair deficiency, originated from germline biallelic MMRD (CMMRD), occurred at young ages, and exhibited ultrahypermutation. PriMMRD2 gliomas were enriched for monoallelic germline MMRD (Lynch syndrome) and older ages. Finally, IDH1 -driven priMMRD3 gliomas harboured germline mutations in MutSalpha ( MSH2 / MSH6 ), lower mutation burden, and distinct imaging patterns. Subgroups also utilized different mechanisms of genomic instability: while priMMRD1 gliomas harboured frequent point mutations in glioma drivers, priMMRD2 and priMMRD3 gliomas had higher rates of copy alteration in the same genes, including CDKN2 A (P<0.0001) and PTE N (P<0.0001). Driver mutations observed in MMRD-driven gliomagenesis could be explained by the specific trinucleotide contexts that compose MMRD mutational signatures. Hotspot mutations in TP53 and IDH1 occurred in contexts commonly mutated by MMRD, while pediatric glioma drivers ( BRAF / H3-3A) , which do not arise from frequently mutated contexts, were rare. Using mutational signatures and variant allele fractions, we built a model of MMRD gliomagenesis where TP53 mutations occur early, followed by secondary mutations in POLE/POLD1 , IDH1 , and other glioma drivers. PriMMRD glioma subgroups differed in their immune microenvironment and response to immunotherapy: priMMRD1 gliomas were associated with the highest 2-year overall survival on anti-PD1 monotherapy (75%), exhibited pro-immune expression signatures, and high CD8+ T-cell infiltration, contrasting priMMRD2 (24%) and priMMRD3 gliomas (0%). These findings suggest that MMRD mutagenesis shapes the unique landscape of priMMRD glioma through driver mutation acquisition and responses to immunotherapy, providing a rational approach for the refinement of subgroup-specific immunotherapy combinations.
利益披露 Disclosure
N. R. Fernandez, None..
Y. Chang, None..
N. M. Nunes, None..
J. R. Dimayacyac, None..
A. Levine, None..
A. Ringel, None..
L. Negm, None..
A. B. Ercan, None..
O. Ahmad, None..
C. Lee, None..
L. Stengs, None..
V. Bianchi, None..
M. Edwards, None..
S. Doherty, None..
J. Chung, None..
L. Nobre, None.
J. Bennett,
Servier Other, Advisory Board.
A. J. Dodgshun, None.
D. T. Jones,
Heidelberg Epignostix GmbH Stock, Founder.
S. M. Pfister,
BioSkryb Other, Honorarium.
Epignostix GmbH Stock.
A. Villani, None..
D. Malkin, None..
V. Ramaswamy, None..
A. Huang, None.
E. Bouffet,
Servier Advisory board member.
M. Aronson, None..
P. B. Dirks, None.
A. Shlien,
NewCode Oncology Co-founder and equity holder.
G. Getz,
IBM ).
Pharmacyclics ).
Bayer ).
Genentech ).
Calico ).
Ultima Genomics ).
Inocras ).
Google ).
Kite ).
Novartis ).
Scorpion Therapeutics Employment, Other Business Ownership, Other, Founder.
Predicta Biosciences Other Business Ownership, Other, Founder.
Antares Other Business Ownership.
Y. E. Maruvka, None..
B. Ertl-Wagner, None.
C. Hawkins,
Servier Honorarium.
A. Das, None..
U. Tabori, None.