©The Author(s) 2018. Published by Baishideng Publishing Group Inc. All rights reserved.
Biomarkers for hepatocellular carcinoma: What’s new on the horizon?
Matthias Ocker, Department of Translational Medicine Oncology, Bayer AG, Berlin 13353, Germany
Matthias Ocker, Charité University Medicine Berlin, Berlin 10117, Germany
Author contributions: Ocker M solely contributed to the manuscript.
Conflict-of-interest statement: The authors have no conflict of interest to declare.
Open-Access: This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Correspondence to: Matthias Ocker, MD, Professor, Department of Translational Medicine Oncology, Bayer AG, Muellerstrasse 178, Berlin 13353, Germany. email@example.com
Telephone: +49-30-468194799 Fax: +49-30-468994799
Received: June 11, 2018
Peer-review started: June 12, 2018
First decision: July 6, 2018
Revised: July 29, 2018
Accepted: August 1, 2018
Article in press: August 1, 2018
Published online: September 21, 2018
Hepatocellular carcinoma (HCC) is the most common primary tumor of the liver and ranks 3rd in cancer-related deaths worldwide[1,2]. While its incidence continues to be high in Africa and Asia, Western countries also showed increasing incidences rates in the past decades due to chronic hepatitis C virus (HCV) infection, alcohol consumption and high rates of obesity linked to non-alcoholic fatty liver disease (NAFLD), non-alcoholic steatohepatitis (NASH) and subsequent development of chronic liver disease with cirrhosis[3,4]. Although, a slight improvement in global HCC mortality was recently reported, certain sub-populations and regions esp. in Western countries still have an unfavourable prognosis and risk assessment with continuous high incidence and mortality[5,6]. Curative treatment options like surgical resection or orthotopic liver transplantation are only feasible in a minority of patients at early disease stages and with preserved liver function. Thus, the overall prognosis for patients with HCC remains unsatisfying, its 5 year survival being a dismal 6.9%, an incidence to mortality ratio of 0.95 and a median overall survival of only 11 mo. For advanced stages, treatment is based on multi-kinase inhibitors like sorafenib or regorafenib, while recent data indicate that immune checkpoint inhibitors will lead to increased response rates also in this setting[8,9].
Biomarkers are defined as characteristics that are measured as indicators of physiologic or pathologic processes or in response to various diagnostic or therapeutic procedures. The reader is referred to the recent definitions of the FDA-NIH Biomarker Working group for full definition of different biomarker types. Various prognostic biomarkers, e.g., α-fetoprotein (AFP), AFP-L3 or Des-γ-carboxyprothrombin (DCP), are currently used or under investigation for the early diagnosis and surveillance of HCC patients. Here, newer biomarkers like osteopontin, glypican-3 or high c-met expression have shown additional value, esp. when combining these parameters as was shown for osteopontin and AFP[11-13]. Yet, little progress was achieved in developing predictive biomarkers for targeted and other novel treatment options[14,15]. In this article, the current status of predictive biomarkers for identification and selection of patients for novel therapies will be discussed.
BIOMARKERS FOR TARGETED THERAPIES
Sorafenib is the current standard of care for advanced HCC. Initial phase 2 data indicated that pretreatment levels of phosphorylated ERK (p-ERK) correlated with time to progression (TTP), which was later confirmed by several preclinical and in vitro studies[17,18]. Due to limited availability of tissue samples, this finding could not be confirmed in the registrational phase 3 study (SHARP trial). Instead, an extensive program investigating different biomarkers from plasma samples was initiated. Surprisingly, none of the investigated biomarkers predicted the response to sorafenib while biomarkers related to clinical performance, e.g., vascular invasion or AFP, as well as markers of angiogenesis like Ang2 or VEGF were shown to be independent predictors of survival in patients with advanced HCC and thus need to be seen rather as prognostic biomarkers. In other smaller studies, elevated p-ERK and VEGF2 tissue levels were shown to be predictive for poor response to sorafenib treatment in a cohort of 77 advanced HCC patients. Similarly, high p-ERK correlated to poor overall survival, but not time to progression, in another study with 44 patients. Interestingly, also the multi-kinase inhibitor regorafenib, which was recently approved for second-line therapy of HCC, stratified patients only on clinical parameters and thus does not have a predictive biomarker available so far for HCC patients, while promising results on plasma circulating cell free DNA were obtained for patients with colorectal cancer.
BIOMARKERS FOR IMMUNE CHECKPOINT INHIBITORS
The recent success of immune checkpoint inhibitors in other cancer diseases also triggered various approaches in HCC. As HCC development is commonly based on chronic inflammatory liver diseases (viral hepatitis, NASH), a clear rationale to investigate this new treatment paradigm is clearly given. Initial results using anti-CTLA-4 or anti-PD-1 antibodies are encouraging and lead to accelerated approval of the anti-PD-1 antibody nivolumab for the treatment of advanced HCC[25,26]. Combination of checkpoint inhibitors seem feasible and the use of such drugs together with locoregional procedures in early disease settings might even further improve outcome of patients.
Still, a significant number of patients (> 60%) do not respond to these novel therapies. Biomarker-based enrichment was initially based on immunohistochemical expression of the respective checkpoint targets, but recent data from various indications suggest that this is not the strongest predictor for treatment response. This could be due to the still tissue based scoring of target expression with an intrinsic risk for sampling error in heterogenous solid tumors[29,30] or due to the still not completely understood biology of checkpoint inhibition as evidenced by approx. 10% of patients who do not express PD-L1 but respond to treatment.
It is intriguing that many promising and potent drugs, e.g., sunitinib, everolimus, brivanib or tivantinib, failed in HCC clinical trials. Besides careful selection of patients based on clinical parameters like liver function or vessel invasion, all-comer trials are nowadays not considered appropriate and identification of specific patient subgroups based on distinct molecular subtypes is therefore urgently requested[33-35].
In HCC, lack of biopsies and different etiologic backgrounds hampered the identification and validation of such markers for the currently available treatments. Even today, practice guidelines do not recommend taking biopsies of every patient although risk of bleeding and needle track seeding are infrequent and should not be seen as a reason against taking a diagnostic biopsy[36,37]. The latest EASL practice guidelines strongly recommend liver biopsy and blood sampling from patients participating in clinical and diagnostic trials.
Genomic profiling established distinct molecular subclasses of HCC that were also linked to specific gene mutations and clinical and histological features. Two major groups, the proliferative (chromosomal instable) and the non-proliferative (chromosomal stable) group, were defined which comprise approx. 50% of HCC cases each. Further analyses defined a stem cell/hepatoblast like and a TGFβ related subgroup in the proliferative group, as well as a hepatocyte like and a Wnt/β-catenin related subgroup in the non-proliferative group[34,38]. While the overall impact of this classification is still under debate, additional common mutations and genetic alterations were described. Overall, mutations in telomerase signaling, the p53 and cell cycle control pathway as well as in Wnt/β-catenin signaling are commonly observed while rarer events include mutations in the Ras/PI3K/mTOR pathway, JAK/STAT signaling and other pathways[38,39]. Interestingly, no clear individual oncogenic driver has been identified in HCC so far and HCC is considered a cancer with medium to low mutational burden.
Analyzing tumor nucleic acids from other sources than tissue, i.e., from circulating tumor cells, cell free DNA/RNA or exosomes, could help to overcome the above mentioned limitations. Liquid biopsies usually detect expression levels, methylation status or mutations of distinct tumor related nucleic acids, including DNA, RNA and miRNAs originating from circulating tumor cells or being shed into the blood directly from living or dying tumor cells. While this approach is considered to reduce sampling error compared to solid tumor biopsies[40,41], there are still technical limitations to this technology. The success rate of detecting circulating tumor cells is depending on the size of the tumor and results seem highly variable, ranging from approx. 25% to 100% success rate within different populations and with different technical approaches. Similar results were obtained for genetic analyses of circulating DNAs[43,44]. Surprisingly, also these studies seem to be underpowered when considering different etiologic and geographical background. A clear advantage of liquid biopsies is also the option of taking serial samples from a patient to detect changes during disease history and imposed by treatment.
PROTEOMICS, GLYCOMICS AND IMAGING
As the liver is the primary secretory and metabolizing organ of the body, the use of proteomics and glycomics (usually by liquid chromatography-mass spectrometry (LC-MS) or matrix assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry analysis) could provide an option to identify novel biomarkers, too, without the current limitations of tissue based analyses. Several proteomic factors, e.g., CD44v9 or Hippocalcin-like 1 (HPCAL1), including various multi-marker approaches were used as prognostic or diagnostic biomarkers for HCC or to predict recurrence[48-52]. Similarly, glycomics-based tests like GlycoCirrhoTest or analysis of N-glycans were developed as further diagnostic tools for better surveillance of patients at risk of HCC development[14,53-55]. So far, these approaches were not used in a predictive setting in advanced HCC.
Beyond LC-MS or MALDI-TOF analysis, functional and molecular imaging represents a further technology to identify potential biomarkers for HCC. Functional imaging is using dynamic computed tomography, dynamic magnetic resonance imaging and diffusion weighted magnetic resonance imaging approaches and is now well established to detect changes in e.g., fibrosis grade or angiogenesis and for early diagnosis of HCC. The development of novel radiotracers (beyond 18F-FDG) for PET imaging could bridge the findings from proteomics and metabolomics analyses to imaging and thus add useful and important information on tissue distribution to these data. Today, molecular imaging for primary liver tumors is still limited by e.g., lack of specific tracer uptake into malignant cells.
Predictive biomarkers are considered key for the success of developing new drugs. The further development of emerging technologies that are not dependent on tissue will also increase our knowledge for the better treatment of patients but more homogeneous study design regarding technologies and patient characteristics need to be done to achieve meaningful sample sizes with results that can be robustly translated into clinical applications.
Manuscript source: Invited manuscript
Specialty type: Gastroenterology and hepatology
Country of origin: Germany
Peer-review report classification
Grade A (Excellent): 0
Grade B (Very good): B
Grade C (Good): C
Grade D (Fair): 0
Grade E (Poor): 0
P- Reviewer: Vradelis S, Yao DF S- Editor: Wang XJ L- Editor: A E- Editor: Yin SY