兒童B細胞急性淋巴性白血病之基因分析

Abstract

B細胞急性淋巴性白血病是最常見的兒童癌症。藉由危險分群導向的治療方案，其治療成績在過去幾十年間已有大幅度的進步。主要的預後相關危險因子包括了基因亞群例如非整倍體 (高二倍體或低二倍體)，融合基因，以及參與B細胞發育相關的基因突變，例如IKZF1及ERG基因等。近幾年來，數個研究團隊利用轉錄體定序對兒童B細胞急性淋巴性白血病進行全面性的研究，發現了數個具有獨特基因表達的新型基因亞群，且註明這些基因亞群也具有預後意義。為了進一步了解在台灣兒童的B細胞急性淋巴性白血病的基因變化，我利用多重連接依賴性探針擴增、DNA指數及反轉錄聚合酶連鎖反應分析常見的基因亞群在台灣的分布率，而未定出基因型的樣本則再以轉錄體定序偵測。藉由轉錄體定序，90%的B細胞急性淋巴性白血病皆可以被分類到特定的基因亞群。藉由與先前所發表的研究進行比較，我也發現了專屬於基因亞群的基因變化，例如在DUX4基因轉位急性淋巴性白血病中發現到高頻率的IKZF1及TP53基因突變，以及在ZNF384基因轉位急性淋巴性白血病發現高頻率的ETV6基因突變。另外也發現到台灣族群的B細胞急性淋巴性白血病基因亞群的分布頻率與先前已發表研究有所差異，例如類費城染色體急性淋巴性白血病在高加索人族群中佔了10-15%，但在台灣族群中卻只有2%。最後我利用多變項分析方法分析，並且確定基因亞群是具有顯著意義的危險因子。藉由本研究的結果，顯示利用轉錄體定序可以檢測B細胞急性淋巴性白血病的基因亞群，且基因亞群的鑑定可以幫助B細胞急性淋巴性白血病的病人進行危險分群。

B-cell acute lymphoblastic leukemia (B-ALL) is the most frequent pediatric malignancy and the outcome had been dramatically improved with risk–directed therapy over the past few decades. The major risk factors include genetic subgroup defined as aneuploidy (high hyperdiploidy or hypodiploidy) or fusion genes and genetic alterations that disrupt B-cell development such as IKZF1 and ERG deletions. Recent years, several large cohort studies had comprehensively analyzed pediatric B-ALL and revealed novel subgroups with distinct gene expression profiles by using RNA-sequencing (RNA-seq) and these subgroups have prognostic significance. To better understand the genomic landscape of pediatric B-ALL in Taiwan, I analyzed recurrent alterations by Multiplex Ligation-dependent Probe Amplification (MLPA), DNA index and RT-PCR. RNA-seq was performed for those cases without known alterations to identify the genetic subgroups. Previous unidentified novel subgroup was discovered by RNA-seq and 90% of cases could be classified as one of the defined genetic subgroups. I also found subgroup-specific alterations that had never been reported such as significant higher rate of IKZF1 deletion and TP53 mutation in DUX4-rearranged ALL and frequent ETV6 alteration in ZNF384-rearranged ALL. The distribution of B-ALL subgroup was different from previous studies such as only 2% of cases were classified as Ph-like in our cohort compared to 10-15% in Caucasian cohorts. Multivariable analysis was performed and subgroup was the significant risk factor. This study showed that RNA-seq can be implemented to discover B-ALL subgroups and identification of genetic subgroup can enhance risk classification of B-ALL.