CDH1
转录组
ARID1A型
癌症研究
基因表达谱
上皮-间质转换
医学
基因
基因签名
基因表达
生物
遗传学
过渡(遗传学)
突变
钙粘蛋白
细胞
作者
Maria Bencivenga,Michele Simbolo,Chiara Ciaparrone,Caterina Vicentini,Lorena Torroni,Maria Liliana Piredda,Michele Sacco,Mariella Alloggio,Claudia Castelli,Anna Tomezzoli,Aldo Scarpa,Giovanni De Manzoni
出处
期刊:Annals of Surgery
[Lippincott Williams & Wilkins]
日期:2022-08-05
卷期号:276 (5): 822-829
被引量:8
标识
DOI:10.1097/sla.0000000000005648
摘要
HYPOTHESIS: Poorly cohesive (PC) gastric cancer (GC) exhibits variable clinical behavior, being extremely aggressive in most cases but more indolent at times. We hypothesized that the integrative genomic and gene expression characterization of a PC GC series could help identifying molecular subtypes with potential clinical implications. MATERIALS AND METHODS: 64 PC GCs were assessed for alterations in 409 genes and 30 cases were subjected to transcriptomic profiling of 20,815 genes. RESULTS: A median of 8.2 mutations per Mb (interquartile range 6.9-10.4) was found and a tumor mutational load >10 muts/Mb was significantly associated with patients' worse survival ( P =0.0024). The most frequent mutated genes were CDH1 and TP53 (each 32.8%) followed by PIK3CA (10.9%). In 15 samples (23.4%), at least 1 chromatin remodeling gene was mutated: KMT2D (5 cases); ARID1A and BAP1 (4 cases each); EZH2 , KMT2A , PBRM1 (1 case each). Eight samples (12.5%) had fusion genes involving CLDN18 gene. Gene expression profiling identified 4 different clusters: cluster A associated with epithelial to mesenchymal transition (EMT) signature; cluster B associated to proliferative signature and EMT; cluster C correlated to hedgehog signaling; cluster D showing no enrichment for any of the previous signatures. Notably, cluster A and B showed a worse prognosis compared with clusters C and D ( P =0.0095). CONCLUSION: integrated genomic and transcriptomic analysis suggest the existence of 4 molecular subtypes of PC GC with prognostic significance where EMT features are associated with a worse outcome.
科研通智能强力驱动
Strongly Powered by AbleSci AI