21Background: Primary liver tissue cancers display consistent increase in global disease burden 22 and mortality. Identification of cell-of-origins for primary liver cancers would be a necessity 23 to expand options for designing relevant therapeutics and preventive medicine for these cancer 24 types. Previous reports on cell-of-origin for primary liver cancers was mainly from animal 25 studies, and integrative research utilizing human specimen data was poorly established.
26Methods: We analyzed a whole-genome sequencing data set for a total of 363 tumor and 27 progenitor tissues along with 423 normal tissue epigenomic marks to predict the cell-of-origin 28 for primary liver cancer subtypes.
29Results: Despite the mixed histological features, the predicted cell-of-origin for mixed 30 hepatocellular carcinoma / intrahepatic cholangiocarcinoma were uniformly predicted as a 31 hepatocytic origin. Individual sample-level prediction revealed differential level of cell-of-32 origin heterogeneity depending on the primary liver cancer types, with more heterogeneity 33 observed in intrahepatic cholangiocarcinomas. Additional analyses on the whole genome 34 sequencing data of hepatic progenitor cells suggest these progenitor cells might not a direct 35 cell-of-origin for liver cancers.
36
Conclusions:These results provide novel insights on the heterogeneous nature and potential 37 contributors of cell-of-origin predictions for primary liver cancers. 38 39 Background 43Primary liver cancers (PLCs) is one of the major cancer types with increasing global disease 44 burden over the years, reaching incidence rates and mortality over 900,000 per year (1, 2). This 45 high morbidity and mortality of PLCs is due to the complex nature of the disease and lacking 46 effective diagnostics and treatment besides multi-kinase inhibitors, thus strongly emphasizing 47 the importance of relevant researches on early diagnosis and extensive drug development. In 48 line with this, several endeavored researches were performed on identifying suitable diagnostic 49 markers and targeted therapy-based treatments for PLCs, including the whole genome and 50 exome-level profiling (3). So far, recent comprehensive efforts on investigating the genomics 51 of PLCs revealed novel insights about the major mutation signatures, sub-classifications, and 52 recurrent somatic mutations in coding regions (TERT, TP53, CTNNB1, KRAS, IDH1/2, etc.) 53 and noncoding regions (NEAT1 and MALAT1), which some of them are driver mutations and 54 may associate with the clinical outcomes (4, 5). More investigations are underway to fully 55 unveil the mechanisms and processes behind the progression of PLCs. 56 One of the complex, unanswered questions associated with the progression of PLCs is the 57 possible cell-of-origins (COOs) corresponding to the various subtypes. PLC not only represents 58 classical hepatocellular carcinoma (HCC) subtype, comprising of ~90% of PLCs, but also 59 includes mixed hepatocellular and cholangiocarcinoma (Mixed) and intrahepatic 60 cholang...