IDH mutations: new genetic signatures in cholangiocarcinoma and therapeutic implications

Keywords::

• 2-hydroxyglutarate
• bile duct
• cholangiocarcinoma
• differentiation
• hypermethylation
• IDH1
• IDH2
• methyltransferase inhibitors
• mutation
• targeted therapy

Identification of IDH1/2 mutations in cholangiocarcinoma

While new molecularly targeted therapy regimens have improved clinical management of cancer patients across a number of disease sites, progress has been slow for applying these to biliary tract cancers. This subset of gastrointestinal malignancies encompasses gallbladder carcinoma and cholangiocarcinoma, which is further subdivided into intrahepatic, perihilar or distal cholangiocarcinoma depending on the site of origin in the biliary tree. Biliary tract cancers afflict approximately 12,000 people in the USA annually and continue to present significant challenges. Due to an incomplete understanding of the risk factors and the lack of screening strategies for most biliary tract cancers, only 10–20% of patients present with early-stage disease that is amenable to curative surgery [1,2]. Prognosis remains poor for patients with locally advanced or metastatic disease, with a median survival of approximately 1 year even with the best combination chemotherapy [3]. Furthermore, the molecular mechanisms that contribute to the pathogenesis of this malignancy remain elusive. This has hampered the ability to identify clinically useful prognostic markers and prominent molecular targets that could advance therapeutic interventions for these patients.

Intrahepatic cholangiocarcinoma represents a significant challenge, as it constitutes the second most common form of liver malignancy. Incidence and mortality rates have also been steadily increasing [1]. Fortunately, our understanding of the underlying disease etiology is expanding. An unexpected observation has recently come from the implementation of broad-based cancer genotyping at our institution (Massachusetts General Hospital, MA, USA), which has been applied across tumor types to clinically guide therapeutic selection based on an underlying mutational signature. When evaluating known site-specific mutations across a panel of 15 cancer genes, recurrent mutations in the gene encoding isocitrate dehydrogenase (IDH)1 were observed in a subset of biliary tract cancer samples [4]. Mutations in this metabolic enzyme were exceedingly rare across the spectrum of other gastrointestinal tumor types tested. An expanded cohort revealed that mutations in the IDH1 or the related IDH2 gene were exclusive to cholangiocarcinomas arising in the liver. The combined IDH1 and IDH2 mutation frequency of 23% was greater than the frequency of other previously identified gene mutations in this malignancy, including PIK3CA, KRAS, NRAS and AKT1. The prominence of this genetic signature suggests a new pathophysiological mechanism in cholangiocarcinoma that may shed light on new therapeutic directions.

Tumorigenic mechanisms of neomorphic IDH1/2 activity

Intrahepatic cholangiocarcinoma is one of a limited number of cancer types, which includes glioma, acute myeloid leukemia (AML) and chondrosarcoma, that have been identified to harbor frequent mutations in IDH1 and IDH2. The diversity of these tumor types suggests that the IDH1/2 mutation plays a role in a fundamental tumorigenic process. In this regard, somatic mutation in IDH1/2 has provided the first example of a cancer mechanism that induces the production of an abnormal metabolic enzyme that can promote metabolic restructuring through a gain-of-function activity. This has caused a paradigm shift in cancer metabolomics and has fostered novel directions in clinical management.

The cytosolic and peroxisomal IDH1 and mitochondrial IDH2 enzymes normally function in a homodimeric complex to catalyze the conversion of isocitrate to α-ketoglutarate. The recurrent cancer mutations are heterozygous and have been confined to conserved residues within the isocitrate binding site of IDH1 (R132) or IDH2 (R140 and R172), while maintaining one respective wild-type allele. The mutant retains the ability to dimerize with the wild-type partner. However, a distinct shift in conformation of the mutant enzyme’s active site impairs its normal enzymatic function and increases its affinity for the products produced locally by the wild-type subunit. This confers neomorphic activity through the reduction of α-ketoglutarate to the metabolite R(–)-2-hydroxyglutarate (2HG), resulting in 2HG accumulation in the tumor cells [5,6]. An excessive concentration of 2HG has been observed in IDH1- and IDH2-mutant cholangiocarcinoma patient samples [4].

The prevailing view is that 2HG functions as an oncometabolite, but its distinct role in tumorigenesis is still being determined. The high degree of structural similarity between 2HG and α-ketoglutarate has been shown to disrupt the normal functioning of multiple α-ketoglutarate-dependent dioxygenases through albeit weak antagonism of α-ketoglutarate cofactor binding [7]. While there are more than 30 dioxygenases that have been identified, not all may be equally susceptible to high concentrations of 2HG, owing partly to different affinities for α-ketoglutarate and/or 2HG. The effects of mutant IDH1 on altering activity of the hypoxia-inducible factor (HIF) prolyl 4-hydroxylases that function to promote proteasomal degradation of the HIF-1α transcriptional subunit are currently being debated [7,8]. However, a role for mutant IDH activity in epigenetic regulation has been identified through the ability of 2HG to significantly remodel the methylome and thus regulate gene expression. 2HG has been found to be a particularly potent inhibitor of a wide range of jumonji domain-containing histone demethylases [7–9]. This has been supported by the ability of in vivo 2HG addition or ectopic mutant IDH1 expression to promote repressive histone methylation patterns that are seen in glioma patient samples [7,10]. Similarly, a transcriptionally repressive DNA hypermethylation phenotype, both globally and at distinct gene promoter CpG islands, has been attributed to mutant IDH1 and IDH2 function in AML and glioma [11–13]. A related mechanistic function of 2HG may be through disruption of the normal activity of the recently discovered TET family of 5-methylcytosine hydroxylases (TET1 and TET2) that are involved in DNA demethylation [7]. Therefore, oncogenic activity of 2HG overproduction in IDH1/2-mutant tumors likely involves aberrant histone and DNA methylation mechanisms that may serve to alter target gene expression.

Clinical implications

The known activities of mutant IDH1 and IDH2 provide potentially optimal targets for therapeutic intervention. Unlike oncogenic mutations that lead to constitutive activation of a normal protein function, IDH1 and IDH2 mutations confer neomorphic activity that is specific to the cancer cell and absent in normal cells. Theoretically, specific targeting of neomorphic IDH activity with small molecule inhibitors could minimize systemic toxic effects that may prove advantageous for use in multi-targeted therapy regimens. These small molecule inhibitors would be particularly relevant for testing in intrahepatic cholangiocarcinoma patients, as nearly a quater of patients were found to harbor IDH1 and IDH2 mutations from our initial observations. A number of pharmaceutical companies are developing IDH inhibitors, and their entry into clinical evaluation is highly anticipated. Therefore, the ability to efficiently identify the most appropriate patients for these initial studies will be increasingly important.

Outside of genetic testing, the excessive accumulation of 2HG may be an attractive biomarker, since it is a direct product of mutant IDH1 or IDH2 activity. Magnetic resonance spectroscopy imaging has recently been reported to be effective in detecting intratumoral levels of 2HG in glioma patients [14], providing a rapid and effective means to evaluate IDH mutational status prior to surgery. While this technology is best suited for tumors such as glioma and possibly chondrosarcoma where body movement can be minimized, it is not clear whether imaging of cholangiocarcinoma in the abdominal cavity would be, at best, restricted to only large tumors. In addition, while mass spectrometry can identify circulating 2HG in the sera of AML patients with absolute accuracy regarding IDH1/2 mutational status [5,6,15], inital studies have demonstrated that this approach may not be useful when evaluating 2HG released from a solid tumor such as glioma [16]. In these respects, DNA-based testing will likely remain the standard approach for identifying mutant IDH1 or IDH2 in cholangiocarcinoma patients.

From a therapeutic standpoint, there is particular appeal in the idea that a single gene mutation can control epigenetic status at the genome level. This may provide the ability to therapeutically reverse silencing across a number of targets simultaneously. It will therefore be of interest to determine whether IDH-targeted therapies can reverse histone and DNA hypermethylation patterns associated with these tumors. Interestingly, hypermethylation has also been frequently observed in IDH wild-type cases. In gliomas, MGMT promoter methylation has been identified in approximately 80% of IDH1-mutant tumors but also in approximately 60% of IDH1-wild-type samples [17]. This suggests that there may be parallel mechanisms outside of mutant IDH activity that can promote a phenotypically similar CpG island methylator phenotype. In support, inactivation of TET2 function through mutation or promoter methylation in AML samples promotes a pattern of DNA hypermethylation that is similar to IDH1/2-mutant samples, suggesting functional redundancy between these two gene families [11,18]. Together, these observations suggest that agents that inhibit DNA methyltransferases (e.g., 5-azacytidine and 5-aza-2´-deoxycytidine) could provide a beneficial treatment strategy in these tumors. Furthermore, histone methylation attributed to mutant IDH1 activity has been shown to block cell differentiation [10]. This raises the question as to whether the incorporation of differentiation therapy with IDH or methyltransferase inhibitors could provide greater efficacy. These approaches may be relevant in cholangiocarcinomas, where CpG island hypermethylation has been reported to be a common epigenetic feature [19,20]. Whether this can be attributed to mutations in IDH1/2, or other related genes that regulate methylation status in cholangiocarcinoma, is an area for further investigation. Until small molecule inhibitors that target IDH mutant activity become available, alternative treatment courses that include methyltransferase inhibitors and/or differentiation therapies may be potential areas of consideration for therapeutic targeting in this malignancy.

Financial & competing interests disclosure

DR Borger is a paid consultant for Bio-Reference Laboratories, Inc. (licensee of SNaPshot). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.