Genomic Analysis of Hepatocellular Carcinoma With Active Hepatitis - PowerPoint PPT Presentation

Genomic Analysis of Hepatocellular Carcinoma With Active Hepatitis B Virus Replication Huat Chye Lim, MD, and John Gordan, MD, PhD Divisions of Hospital Medicine and Hematology/Oncology University of California, San Francisco 13 th Annual Conference 20 ► 22 September 2019 13 th ILCA Annual Conference Chicago, USA 20 ► 22 September 2019 │ Chicago, USA

Introduction Synopsis HCC with active HBV replication represents a molecularly distinct subset of HCC associated with differences in mutations, gene expression and survival. Objectives for this talk 1. Which genes are differentially mutated in HCC with active HBV replication? 2. Which genes are differentially expressed in HCC with active HBV replication? 3. How does HBV replication status affect survival in HCC? Background: HBV and HBV-related HCC • Small enveloped DNA virus with partially double stranded genome • Persists in nuclei of infected hepatocytes via episomal cccDNA • Rarely integrates into host genome • HBV-related HCC known to be associated with clinical and genomic differences 13 th ILCA Annual Conference 20 ► 22 September 2019 │ Chicago, USA

Methods • We used GATK PathSeq software, which All tumor RNA-Seq reads performs sequential computational subtraction, to measure HBV RNA in HCC Quality filtering and duplicate removal tumors • Tumor RNA-Seq data were obtained from Low quality and Filtered reads duplicate reads two databases: o TCGA – n = 371 Subtraction of human reads o ICGC – n = 68 (from LIRI-JP project) Non- Low quality and human Human reads • We classified tumors as HBV RNA+ if more duplicate reads reads than 1 HBV RNA read was detected per Mapping to known microbes million human reads Other • We investigated association between HBV Low quality and microbe Human reads duplicate reads RNA status and nonsynonymous somatic reads mutations, gene set expression and survival HBV reads 13 th ILCA Annual Conference Reference: Walker MA, et al. Bioinformatics. 2018;34(24):4287-4289. 20 ► 22 September 2019 │ Chicago, USA

Results: HBV RNA Status is Associated With Differences in Clinical Characteristics HBV RNA+ HBV RNA- p Total patients 124 315 N/A • HBV RNA+ status was TCGA 100 81% 271 86% Cohort N/A associated with: ICGC (LIRI-JP) 24 19% 44 14% o Male gender Male 100 81% 201 64% Gender 0.0006 (*) Female 24 19% 114 36% o Younger age Mean ± SD 54.2 ± 12.0 62.9 ± 13.1 Age at diagnosis < 0.0001 (**) o Higher grade HBV clinical history 87 70% 44 14% < 0.0001 (*) Risk factors HCV clinical history 2 2% 79 25% < 0.0001 (*) o HBV history Alcohol consumption 45 36% 106 34% NS o No HCV history Grade I 12 10% 47 15% • There was no Grade II 50 40% 161 51% Edmondson grade Grade III 51 41% 83 26% 0.0003 (*) association with stage, at diagnosis Grade IV 8 6% 4 1% vascular invasion or Unknown 3 2% 20 6% alcohol consumption Stage I 54 44% 129 41% Stage II 28 23% 84 27% p values from χ 2 (*) and Mann- • Pathologic stage Stage III 38 31% 73 23% NS Whitney (**) tests at diagnosis • Stage IV 3 2% 10 3% Stage determined using either AJCC (for TCGA) or LCSGJ (for Unknown 1 1% 19 6% ICGC LIRI-JP) criteria Present 37 30% 105 33% Vascular invasion Absent 65 52% 176 56% NS 13 th ILCA Annual Conference Unknown 22 18% 34 11% 20 ► 22 September 2019 │ Chicago, USA

Results: HBV RNA Status is Associated With Differential Gene Mutation Rates 0,08 TP53 • Figure shows the 94 genes where 0,07 DNAH6 Preferentially mutated in HBV RNA+ tumors nonsynonymous mutation rate BAP1 Preferentially mutated in HBV RNA- tumors depended significantly on HBV 0,06 SYNE2 RNA status FBN1 HBV RNA- Mutation Rate 0,05 • Most (82/94) were preferentially SVEP1 mutated in HBV RNA+ tumors CFAP47 0,04 • BPTF Bolded genes are among the 30 BRCA2, CHD9, significantly-mutated genes DOPEY2, TNXB, 0,03 TRIP12 identified in 2017 TCGA HCC Cell NBEA COL4A5, ENGASE, paper TDRD5 ZNF208 ZNF135 0,02 BRD7 CSDE1, EXO1 o TP53 and CDKN2A were more CC2D2A, EGFLAM, SSPO NRXN1 frequently mutated in HBV RNA+ FBXO42, NCOA2, SPTBN2 0,01 CHD5 OR5D14 o CDKN2A BAP1 was more frequently CDH4 and 18 other CNOT2 genes mutated in HBV RNA- 0,00 MED12 MAPK9 and 36 other o TP53 had substantially higher CCNL2, FOXG1, genes KAT6A, NBEA mutation rates than other genes -0,01 -0,01 0,01 0,03 0,05 0,07 0,09 0.35 0,11 • Circle size is inversely proportional to log( p ) HBV RNA+ Mutation Rate • Bolded genes are TCGA top 30 SMGs 13 th ILCA Annual Conference Reference: Cancer Genome Atlas Research Network. Cell. 2017;169(7):1327-1341.e23. 20 ► 22 September 2019 │ Chicago, USA

Results: HBV RNA+ Status is Associated With Differential Gene Set Expression Cell cycle regulation Chromatin • Figure shows GSEA enrichment map of Gene Mitotic modification spindle Ontology gene sets with FDR < 10% in TCGA regulation dataset ( n = 531 of 4,464) DNA • replication All such gene sets were enriched only in HBV RNA+ tumors Genes upregulated in FDR q • Enriched gene sets included: Boyault subclass G1-G3 HCC 0.003 Methylation o Multiple DNA damage repair pathways and Hoshida subclass S2 HCC 0.011 DNA o Genes upregulated by MYC and mTORC1 chromatin damage modification Genes upregulated in “proliferative” HCC: Lee subclass A HCC 0.012 o repair ▪ Boyault subclass G1-G3 HCC Chiang “proliferation” subclass HCC 0.015 ▪ Hoshida subclass S2 HCC Nucleases Nuclear RNA ▪ Lee subclass A HCC transport processing ▪ Chiang “proliferation” subclass HCC Enrichment in HBV RNA+ Translation • We then evaluated association between HBV 0.0001 FDR 0.1 RNA status and several measures of genomic Circle size is proportional RNA processing instability to gene set size and splicing Transcription 13 th ILCA Annual Conference References: Boyault S, et al. Hepatology. 2007;45(1):42-52. Hoshida Y, et al. Cancer Res. 2009;69(18):7385-92. Lee JS, et al. Hepatology. 2004;40(3):667-76. Chiang DY, et al. Cancer Res. 2008;68(16):6779-88. 20 ► 22 September 2019 │ Chicago, USA

Results: HBV RNA+ Status is Associated With Increased Homologous Recombination Deficiency (HRD) Score p = 1e-6 • HRD score for TCGA dataset was calculated as the sum of three 60 independent HRD measures: o Large-scale state transitions o Loss of heterozygosity HRD score o Telomeric allelic imbalance 40 • HBV RNA+ status was associated with increased HRD score (22.19 for RNA+ vs. 15.97 for RNA-, p = 1e-6) 20 o There was no association with tumor mutational burden (TMB) • Error bars show mean ± SD • p value from Mann-Whitney test 0 HBV RNA+ HBV RNA- 13 th ILCA Annual Conference Reference: Knijnenburg TA, et al. Cell Rep. 2018;23(1):239-254.e6. 20 ► 22 September 2019 │ Chicago, USA

Results: HBV RNA Status is Associated With Survival Differences HBV RNA+ HBV RNA- All • BAP1 mutations were associated with We used Cox multivariate Cox Cox Cox Gene p p p increased survival in HBV RNA- patients regression to evaluate effect of Coefficient Coefficient Coefficient TP53 0.0420 0.9148 0.5393 0.0284 0.4334 0.0356 mutation and HBV RNA status 100 CTNNB1 -0.1966 0.6505 0.1321 0.6242 0.0087 0.9680 on overall survival ALB -0.4302 0.5601 -0.4065 0.2834 -0.3447 0.3030 Percent survival o Covariates: HBV RNA status, AXIN1 0.5324 0.3328 0.3711 0.3730 0.3568 0.2738 BAP1 BAP1 -15.3500 0.9979 -1.9399 0.0081 -1.7160 0.0183 mutation status, age, sex, 50 KEAP1 1.4484 0.0135 -0.4677 0.5147 0.1392 0.7420 KEAP1 grade, stage, cohort NFE2L2 -16.2300 0.9980 0.3430 0.4209 0.2735 0.5197 • Table shows Cox coefficients for LZTR1 -17.5800 0.9977 0.0118 0.9908 -0.6234 0.5365 BAP1+ p = 0.008 BAP1- RB1 1.4337 0.0428 0.3752 0.4280 0.4903 0.1877 the 30 SMGs identified in 2017 0 PIK3CA -17.2800 0.9965 0.1847 0.7583 -0.2901 0.6274 0 1000 2000 3000 4000 TCGA HCC Cell paper RPS6KA3 -16.4300 0.9976 0.0003 0.9996 -0.1785 0.7270 Time (days) AZIN1 0.5137 0.6240 -13.8100 0.9958 0.9450 0.3518 • Mutations associated with KRAS -15.3500 0.9979 -0.0901 0.9288 -0.4257 0.6728 survival difference in HBV RNA-: KEAP1 mutations were associated with IL6ST -15.7900 0.9971 -1.0137 0.1645 -1.0452 0.1492 RP1L1 -15.6800 0.9970 0.5156 0.3868 0.4072 0.4892 o Increased survival: BAP1 decreased survival in HBV RNA+ patients CDKN2A -1.0873 0.3009 -16.0700 0.9956 -1.0655 0.2930 o Decreased survival: TP53, EEF1A1 NA NA 0.8659 0.0447 0.9484 0.0276 100 EEF1A1, ARID1A, APOB ARID2 1.2225 0.0462 0.2595 0.5757 0.6644 0.0464 ARID1A 0.5161 0.4150 1.0912 0.0008 0.9356 0.0011 Percent survival • Mutations associated with GPATCH4 1.7634 0.0952 -15.4800 0.9948 -0.0130 0.9900 survival difference in HBV RNA+: ACVR2A -0.3718 0.7198 0.4871 0.3541 0.3774 0.4150 50 APOB -1.5385 0.1443 0.6915 0.0164 0.1962 0.4749 o Decreased survival: KEAP1, CREB3L3 NA NA 5.5943 0.0001 6.0263 0.0000 KEAP1+ RB1, ARID2 NRAS -15.5300 0.9978 -14.9700 0.9959 -14.9500 0.9950 p = 0.01 KEAP1- AHCTF1 -14.6900 0.9971 -0.0716 0.8900 -0.0266 0.9587 0 0 1000 2000 3000 HIST1H1C -16.0800 0.9978 -0.4665 0.6444 -0.9352 0.3529 13 th ILCA Annual Conference Time (days) Reference: Cancer Genome Atlas Research Network. Cell. 2017;169(7):1327-1341.e23. 20 ► 22 September 2019 │ Chicago, USA