Genetic factors in the clinical predictive model for hepatocellular carcinoma: Evidence from genetic association analyses

Integrating genetic variants into clinical models for hepatocellular carcinoma risk stratification in cirrhosisJournal of HepatologyVol. 78Issue 3PreviewIdentifying individuals at higher risk of developing hepatocellular carcinoma (HCC) is pivotal to improve the performance of surveillance strategies. Herein, we aimed to evaluate the ability of single nucleotide polymorphisms (SNPs) to refine HCC risk stratification. Full-Text PDF Reply to: “Genetic factors in the clinical predictive model for hepatocellular carcinoma: Evidence from genetic association analyses”Journal of HepatologyVol. 79Issue 1PreviewWe read with great interest the comment of Chen et al.1 to our recent contribution to the Journal of Hepatology.2 In their letter, the authors highlight that genetic variants associated with hepatocellular carcinoma (HCC) differ as a function of European or East Asian ancestry. Their opinion is supported by the analysis of previously published data from genome-wide association studies (GWASs) conducted in Asian3 and European4 populations (the latter being performed by our consortium in France and Belgium); they found that only PNPLA3 rs738409 was associated with HCC in both. Full-Text PDF We read with great interest the study by Pierre Nahon et al. where the traditional clinical model was improved to better predict the probability of hepatocellular carcinoma (HCC) development in patients with cirrhosis, by incorporating information on seven single nucleotide polymorphisms (SNPs), including rs738409 (PNPLA3), rs58542926 (TM6SF2), rs187429064 (TM6SF2), rs641738 (MBOAT7), rs72613567 (HSD17B13), rs429358 (APOE) and rs708113 (WNT3A-WNT9A).[1]Nahon P. Bamba-Funck J. Layese R. Trépo E. Zucman-Rossi J. Cagnot C. et al.Integrating genetic variants into clinical models for hepatocellular carcinoma risk stratification in cirrhosis.J Hepatol. 2022;78:584-595; Abstract Full Text Full Text PDF Scopus (7) Google Scholar This study is thought-provoking and should have great potential application in clinical practice. However, this model’s performance is modestly improved as its C index is less than 0.8 and whether such a model is suitable for other populations is still unknown.[1]Nahon P. Bamba-Funck J. Layese R. Trépo E. Zucman-Rossi J. Cagnot C. et al.Integrating genetic variants into clinical models for hepatocellular carcinoma risk stratification in cirrhosis.J Hepatol. 2022;78:584-595; Abstract Full Text Full Text PDF Scopus (7) Google Scholar We think that this model could be greatly improved and applied to other ancestries if taking the genetic effects in other populations into consideration.[2]Graham S.E. Clarke S.L. Wu K.H. Kanoni S. Zajac G.J.M. Ramdas S. et al.The power of genetic diversity in genome-wide association studies of lipids.Nature. 2021; 600: 675-679Crossref PubMed Scopus (113) Google Scholar Besides, this model should be enhanced if considering the genetic risk score from other correlated traits.[3]Krapohl E. Patel H. Newhouse S. Curtis C.J. von Stumm S. Dale P.S. et al.Multi-polygenic score approach to trait prediction.Mol Psychiatry. 2018; 23: 1368-1374Crossref PubMed Scopus (99) Google Scholar It is essential to appraise the genetic effects on HCC before predicting the risk of transition from cirrhosis to carcinoma. We retrieved the full summary statistics of two genome-wide association studies (GWAS) for HCC. The GWAS of East Asians consisted of 1,866 cases (1,384 males and 482 females) and 195,745 controls (97,655 males and 98,090 females), adjusting for sex, age and the first five principal components using a generalized linear mixed model provided by SAIGE software.[4]Ishigaki K. Akiyama M. Kanai M. Takahashi A. Kawakami E. Sugishita H. et al.Large-scale genome-wide association study in a Japanese population identifies novel susceptibility loci across different diseases.Nat Genet. 2020; 52: 669-679Crossref PubMed Scopus (176) Google Scholar,[5]Zhou W. Nielsen J.B. Fritsche L.G. Dey R. Gabrielsen M.E. Wolford B.N. et al.Efficiently controlling for case-control imbalance and sample relatedness in large-scale genetic association studies.Nat Genet. 2018; 50: 1335-1341Crossref PubMed Scopus (449) Google Scholar The GWAS of Europeans included 775 cases (698 males and 77 females) and 1,332 controls (962 males and 370 females), adjusting for sex, age, the first 10 principal components and fibrosis stage using a logistic regression model provided by PLINK software.[6]Trépo E. Caruso S. Yang J. Imbeaud S. Couchy G. Bayard Q. et al.Common genetic variation in alcohol-related hepatocellular carcinoma: a case-control genome-wide association study.Lancet Oncol. 2022; 23: 161-171Abstract Full Text Full Text PDF PubMed Scopus (21) Google Scholar,[7]Purcell S. Neale B. Todd-Brown K. Thomas L. Ferreira M.A. Bender D. et al.PLINK: a tool set for whole-genome association and population-based linkage analyses.Am J Hum Genet. 2007; 81: 559-575Abstract Full Text Full Text PDF PubMed Scopus (20664) Google Scholar The HCC definition of East Asians refers to all-cause HCC while that of Europeans only refers to alcohol-related HCC. Initially, we compared the four SNPs’ (rs429358, rs58542926, rs708113, and rs738409) genetic associations with HCC in both ancestries as information on the other three SNPs was missing in either GWAS. All four SNPs are significantly associated with the risk of HCC in Europeans while none of them, except for rs738409, is associated with HCC risk in East Asians (Fig. 1A). In addition, the SNP rs429358 is merely associated with HCC risk at a low significance level in Europeans (p value = 0.033), indicating that it should have relatively lower power in predicting HCC risk and should be replaced by other genetic variants. Furthermore, we compared the different associations of genetic loci with HCC risk in both ancestries and displayed the results using the Miami plot (Fig. 1B). A clear difference was observed in chromosome 6. The most significant SNP in chromosome 6 of East Asians is rs200715955, and we currently know little about this genetic locus. Although SNPs in chromosome 19 increase HCC risk in both ancestries, the specific loci are different. For example, the top SNP rs8107030 in chromosome 19 of East Asians is located in the IFNL4 gene region, suggesting the immune mechanism should be considered. The top SNP in chromosome 19 of Europeans is rs739846 (SUGP1). Such discrepancies might be caused by ethnic specificity and different HCC subtypes. Anyway, these results implicated that the predictive model might not be extrapolated to East Asians if using the current SNPs and that the inclusion of additional genetic variants in other genes, such as IFNL4 and SUGP1, might improve predictive power. Generally, the current predictive model can be improved from two aspects. On one hand, using genetic variants with the incorporation of genetic effects from multi-ancestry GWAS should improve the generalizability and predictive power of the current model since the performance of polygenic prediction can be substantially improved by increasing population diversity rather than studying additional European individuals.[2]Graham S.E. Clarke S.L. Wu K.H. Kanoni S. Zajac G.J.M. Ramdas S. et al.The power of genetic diversity in genome-wide association studies of lipids.Nature. 2021; 600: 675-679Crossref PubMed Scopus (113) Google Scholar On the other hand, using the genetic risk score constructed by other correlated traits and additional genes might explain more variance in HCC phenotypes. Notably, HCC can be caused by viral hepatitis, alcohol, and metabolic cofactors.[8]Vogel A. Meyer T. Sapisochin G. Salem R. Saborowski A. Hepatocellular carcinoma.Lancet. 2022; 400: 1345-1362Abstract Full Text Full Text PDF PubMed Scopus (117) Google Scholar The genetic susceptibility of HCC subtypes can be consolidated by a multi-trait genetic risk score, which can be obtained from the GWAS of these subtypes. Alternatively, the consolidated multi-trait genetic risk score can be replaced by a significant pathway shared by these subtypes if discovered. The latter might be much more difficult to achieve but can reduce costs in clinical practice. Besides, the SNPs used in the current model are mainly associated with NAFLD and lipid metabolism, however, the up-to-date multi-omics analysis of non-alcoholic fatty liver disease discovered that two mutations (p.His48Arg in ADH1B and p.Cys282Tyr in HFE) were not proportional to the effect on proton-density fat fraction, suggesting that the underlying genetic effects associated with other traits should be used to construct a better predictive model.[9]Sveinbjornsson G. Ulfarsson M.O. Thorolfsdottir R.B. Jonsson B.A. Einarsson E. Gunnlaugsson G. et al.Multiomics study of nonalcoholic fatty liver disease.Nat Genet. 2022; 54: 1652-1663Crossref PubMed Scopus (17) Google Scholar We appreciate the scientific work by Pierre Nahon et al. and hope that more patients worldwide can benefit from such predictive models. This study received no funding from any organization or government. No potential conflicts of interest should be disclosed in this study. Please refer to the accompanying ICMJE disclosure forms for further details. L.C. acquired the data, performed the genetic analyses, and drafted the manuscript. Z.F. revised the manuscript. Y.Z. and H.Y. checked the integrity of the data analysis. G.L. proposed the idea, revised the manuscript, and was responsible for the integrity of data acquisition and statistical analyses. The GWAS summary statistics for HCC from Biobank Japan can be accessed via http://jenger.riken.jp/. The GWAS summary statistics for HCC of Europeans can be accessed via https://www.ebi.ac.uk/gwas/studies/GCST90092003. We would like to thank all the other investigators for making summary statistics openly available. The following are the supplementary data to this article: Download .pdf (.53 MB) Help with pdf files Multimedia component 1