Mutational landscape of HCC—the end of the beginning

Abstract

Next-generation sequencing analysis and characterization of the microenvironment ‘field-effect’ that promotes hepatocellular carcinoma (HCC) development has revealed critical players and potential targets for chemoprevention. A biomarker-based drug development strategy is needed to improve future HCC clinical trials and therapies.

Liver cancer is the sixteenth global cause of death and the second cause of cancer death after lung cancer.¹ The incidence and mortality from this cancer are increasing worldwide.¹ Hepatocellular carcinoma (HCC)—the most frequent liver cancer affecting around 700,000 patients every year—occurs in patients with underlying chronic liver disease, a feature that has direct implications in the pathogenesis and management of this neoplasm.

High-resolution analysis of molecular alterations in human malignancies has become a research priority. Large-scale mutational screening approaches have enabled the identification of new disease drivers in some solid tumours such as lung, breast or melanoma. Unfortunately, liver cancer has not reached the point of molecular-based treatment stratification. Nonetheless, recent studies have provided a broad picture of the mutational profile in HCC and identified an average of 30–40 mutations per tumour, few of which are expected to be driver mutations.^2–4 Three independent deep-sequencing studies confirmed TP53 and CTNNB1 (which encodes for β-catenin) as frequently mutated in HCC. Mutations in these genes are frequently mutually exclusive.^2,3 In addition, these studies discovered novel mutations associated with HCC in different members of the chromatin remodelling pathway (ARID1A and ARID2), in genes involved in ubiquitination (KEAP1), in RAS/MAPK signalling (RPS6KA3) and in oxidative stress (NFE2L2). These results allowed Guichard and colleagues³ to group the most relevant molecular alterations in HCC in five major signalling cascades: Wnt signalling, TP53 signalling, Ras signalling, oxidative stress, and chromatin remodelling. Recent studies using whole-genome sequencing have also identified mutations in JAK1 in 9% of hepatitis B virus (HBV)-related HCC.⁴ JAK1, a member of the Janus tyrosine kinase family, has a role in immunity, cell growth and differentiation, mostly via STAT signalling, and mutations in JAK1 have been associated with some types of leukaemia. Functional validation of these HCC-related JAK1 mutations in experimental models suggested that JAK1 inhibition represents an attractive new therapeutic target. It was also clear that genes such as EGFR, BRAF, PIK3CA or KRAS, commonly mutated in other solid tumours, are rarely mutated in HCC (<5% of cases; Table 1).^2–4 Several genomic studies have contributed to the identification of molecular subclasses of HCC. As predicted, recurring mutations in HCC fall within certain gene-expression patterns.⁵ For example, CTNNB1 mutations are significantly enriched in patients within the WNT molecular subclass as defined by using unsupervised microarray expression data.⁵

Table 1.

Mutation portrait of human hepatocellular carcinoma

Gene	Pathway/gene function involved	Estimated frequency‡ (%)
Genes frequently mutated in HCC
TERT promoter*	Telomere stability	60
TP53	Genome integrity	20–30
CTNNB1	WNT signalling	15–25
ARID1A	Chromatin remodelling	10–16
TTN	Chromosome segregation	4–10
NFE2L2	Oxidative stress	6–10
JAKI	JAK/STAT signalling	0–9
AXIN1	WNT signalling	4–9
ARID2	Chromatin remodelling	5–7
KEAP1	Ubiquitination	3–8
Genes frequently mutated in other solid tumours, but rarely muted in HCC
IDH1, IDH2	NAPDH metabolism	<5
EGFR	Growth factor signalling	<5
BRAF	RAS/MAPK signalling	<5
KRAS, NRAS	RAS/MAPK signalling	<5
PIK3CA	AKT signalling	<5
PTEN	AKT signalling	<5

^*TERT mutation frequency based on targeted sequencing.⁶

^‡Based on deep-sequencing studies.

In 2013, mutations in the promoter of the telomerase reverse-transcriptase (TERT) gene have emerged as the most prevalent somatic mutation affecting 60% of HCC cases,⁶ an important finding that has impacted the landscape of genetic alterations in HCC. These mutations created a potential binding site for the E-twenty six (ETS) transcription factor and are predicted to increase TERT responsive transcription. Interestingly, mutations within the TERT promoter were also present in 25% of pre-neoplastic cirrhotic macronodules and 44% of adenomas with malignant transformation. A potential role for TERT deregulation in hepatocarcinogenesis has been previously postulated, but the identification of mutations in the TERT promoter provides mechanistic insights, establishing a link between TERT alteration and functional deregulation. In fact, Nault et al.⁶ showed a correlation between mutations in TERT promoter and TERT transcript expression in preneoplastic lesions in hepatocarcinogenesis. Furthermore, the high prevalence of these mutations in preneoplastic lesions points to TERT as the first ‘gatekeeper’ gene in HCC, opening new therapeutic routes in the chemoprevention arena.

In terms of changes in clinical decision-making in HCC management, not much progress was made in 2013. After the approval of sorafenib, six randomized controlled trials that could have potentially changed the standard of care for patients with HCC, assessing drugs that included brivanib (a FGFR, VEGFR and PDGFR inhibitor), erlotinib (EGFR inhibitor), linifanib (dual inhibitor of VEGFR and PDGFR) and everolimus (mTOR inhibitor) have reported negative findings in the first and second-line settings.^7,8 The phase III study testing brivanib versus sorafenib in the first-line treatment of more than 1,000 patients with advanced-stage HCC did not meet the primary end point of improvement in overall survival with a non-inferiority design.⁷ Similarly, the phase III study testing brivanib compared with placebo in the second-line setting after disease progression following sorafenib treatment failed to demonstrate survival benefits for brivanib.⁸ These negative results highlight the intrinsic resistance of HCC to therapies, the specific toxicity profile of patients with underlying liver diseases, and the complexity of trial design in this heterogeneous cancer. Of note, none of these studies selected patients based on predicted target deregulation, a paradigm that proved to be remarkably effective in other solid tumours. Therefore, a potential beneficial effect in a subset of patients could have been missed as a result of treating the overall population. Luckily, a change in this paradigm is starting to permeate drug development in HCC. A phase III trial testing tivantinib in HCC patients whose tumours have a high-MET expression and a phase II proof-of-concept trial evaluating the MEK inhibitor, refametinib, in patients with tumours harbouring RAS mutations are ongoing.

When considering the global impact of liver cancer, prevention is the optimal approach. This was clearly demonstrated years ago when a direct correlation between nationwide HBV vaccination programmes and decreased liver cancer incidence was reported in Taiwan. Antiviral HBV and HCV therapies have been able to decrease and sometimes abrogate the transition from chronic hepatitis to cirrhosis, and thus decrease the incidence of HCC. Specific efforts are still needed to prevent liver cancer development in patients at high risk—mainly those patients with cirrhosis, who represent approximately 1% of the population—as one third of them will develop HCC in their lifetime. A 186-gene signature is now able to identify the 20% of cirrhotic patients with the highest risk of developing cancer or who will die due to progressive liver dysfunction.⁹ This ‘field-effect’ gene signature is enriched in genes involved in inflammation, oxidative stress and cell proliferation (such as IL-6 and EGF), and might enable patient selection for chemoprevention trials.⁹

A recent study has likewise elegantly shown the molecular interactions between dysplastic nodules and the microenvironment that resulted in the development of liver tumours in experimental models.¹⁰ In this study, the researchers found a pivotal role of aberrant IL-6 expression in malignant transformation, mostly as a consequence of the deregulation of a non-coding RNA (LIN28). The researchers isolated HCC progenitor cells from different mouse models, and showed that autocrine IL-6 secretion promotes their malignant transformation.¹⁰ Interestingly, the study presents functional evidence on how the malignant potential of these progenitor cells is highly dependent on their microenvironment, further emphasizing the role of IL-6 signalling as a potential chemoprevention target.

We can, therefore, state that liver cancer has reached ‘the end of beginning’ in terms of genome characterization. There is an overall picture of the mutation rate and distribution, but whether these events have a driving role in HCC progression or if they define oncogenic addiction loops in human HCC is still unclear. Results from negative phase III trials in all-comers reinforce the concept that molecular information should be incorporated in the drug development process for liver cancer. Biomarker-driven strategies to known targets or deregulated pathways should be encouraged and represent the future for clinical trial design in this field.

Key advances.

• The most prevalent mutations in HCC affect the TERT promoter, TP53 and CTNNB1. Few ‘druggable’ mutations have been reported^4,6

• IL-6 is emerging as a reliable target for chemoprevention strategies in HCC^9,10

• After the approval of sorafenib, six pivotal phase III trials reported negative findings^7,8