Hepatocellular carcinoma (HCC) is the most common type of liver cancer. With more than 900,000 new cases, and over 830,000 deaths worldwide, incidence and mortality rates continue to rise (CA Cancer J Clin 2021; https://doi.org/10.3322/caac.21660).
Most HCC cancers are found when the disease is already metastatic, making successful treatment unlikely. Like all cancers, early detection raises the chances to improve cancer-related mortality. Increasingly, studies are showing that cancer development is influenced by interactions between viruses and the immune system. One of the major risk factors for HCC is viral infection, typically hepatitis B (HBV) and hepatitis C (HCV), and next-generation sequencing technologies have revealed an association of other viruses with HCC, suggesting that a range of viruses may possibly be involved in the pathogenesis of HCC (Gigascience 2018; doi: 10.1093/gigascience/giy135).
The current methods for screening people with known HCC risk factors are biannual screening with ultrasound with or without a blood test for alpha-fetoprotein (AFP). However, the risk greatly varies among individuals, and the current strategy for early detection and surveillance remains suboptimal. Consequently, novel strategies are urgently needed for the early detection and diagnosis of people who have the highest risk for HCC with the goal of improving outcomes.
In a recent study published in Cell, researchers at the National Cancer Institute (NCI) took a novel approach and explored whether measuring people's lifetime exposure to viruses could help predict their risk of developing HCC (2020; https://doi.org/10.1016/j.cell.2020.05.038). The study also involved researchers from the National Institute of Diabetes and Digestive and Kidney Diseases and several academic centers.
Led by Xin Wei Wang, PhD, from the NCI, the research team reasoned that unique viral exposure signatures (VESs) ensuing from virus-host interactions could reflect a cascade of events that may change the risk of developing HCC. "Such signatures may serve as early detection biomarkers and offer knowledge about potentially modifiable factors for early onset of HCC," the authors write.
To explore the possibility, the team developed a new HCC screening blood test using a synthetic human virome, called VirScan. It probes for a viral signature of past exposures to all known human viruses by applying a phage display library that covers 93,904 viral epitopes, representing 206 human viral species and over 1,000 viral strains.
In this study, the researchers profiled serological samples from 899 people currently enrolled in an NCI-University of Maryland (NCI-UMD) case-control study of liver cancer, including 150 who had HCC, to detect the exposure history of these individuals. Using this high-throughput method, researchers developed the unique VES that could discriminate between individuals with HCC compared with people with chronic liver disease and healthy volunteers. They identified a "viral signature" of exposure to 61 viruses that differed substantially between people diagnosed with HCC and those without the disease.
The researchers then tested the signature in a longitudinal cohort with 173 at-risk patients who were part of a 20-year study. During that time, 44 of the participants developed HCC.
The team discovered that, among at-risk individuals in the validation cohort, the viral exposure signature correctly identified those who developed HCC (area under the curve, AUC=0.98 [95% CI 0.97-1]). An AUC of 0.5 indicates that a test is no better than chance in identifying disease, whereas an AUC of 1.0 represents a test with perfect accuracy. Importantly, the screen also detected HCC in their blood samples taken at the start of the study up to 10 years before diagnosis (AUC=0.91 [95% CI 0.87-0.96]). Moreover, the viral signature was able to distinguish which people would go on to develop HCC much more accurately than the AFP test (AUC=0.91 vs. 0.62).
Notably, by comparing VirScan results as predicted values of HCV and HBV status against those from medical charts of the NCI-UMD cohort as true values, the researchers found that VirScan demonstrated superior accuracy (0.73; 95% CI, 0.67-0.79), sensitivity (0.84; 95% CI, 0.77-0.89), and positive predictive values (0.80; 95% CI, 0.76-0.84) for HCV than the accuracy (0.48; 95% CI, 0.41-0.54), sensitivity (0.48; 95% CI, 0.40-0.57), and positive predictive values (0.55; 95% CI, 0.49-0.61) for HBV.
The researchers conclude that these findings demonstrate that VirScan can predict HCC risk before clinical diagnosis among at-risk patients. Next, the team is planning to test their viral signature in a new study of people at high risk of HCC who have not yet developed the disease.
Xin Wei Wang, PhD, senior investigator and Deputy Chief of the Laboratory of Human Carcinogenesis; and Co-Director of the Liver Cancer Program at the NCI Center for Cancer Research, provided further insights into their study.
Oncology Times: This study establishes a viral exposure signature that can predict HCC among at-risk patients prior to a clinical diagnosis. What motivated you to pursue this research study?
Wang: "Traditionally, liver cancer diagnosis relies on image and circulating tumor cell markers such as alpha-fetoprotein. Detecting features of cancer cells has been the main principle for most if not all diagnostic strategies for cancers. Cancer surveillance programs have been implemented for decades for at-risk individuals for the development of HCC. However, these practices continue to yield mixed results related to the effectiveness in detecting HCC at an early stage or to providing survival benefit. Notably, a majority of HCC patients are still diagnosed at an advanced stage, which precludes their chance to receive potentially curative therapies, leading to poor survival.
"We clearly need a paradigm shift in liver cancer diagnosis. We know that, while cancer is a genetic disease, molecular features of cancer cells may change over time and not all cancer cells in a tumor lesion have the same characteristics. We also know that cancer may be due to a failure of immunosurveillance. Thus, we hypothesize that a history of viral exposure reflecting from virus-host interaction may serve as a window of host immunity and thus stable biomarkers for early onset of liver cancer. While this study is only applicable for HCC, we think the same strategy may be useful for other cancer types."
Oncology Times: What were some of the factors which posed challenges in the development of this novel synthetic virome technology?
Wang: "While this technology is fairly sensitive to detect history of viral exposure in general, its sensitivity and specificity for a specific virus such as HBV are suboptimal. The approach is also sensitive to day-to-day variations, which could be overcome by implementing internal controls. The current phage library consists of about 100,000 unique phage-displayed viral epitopes. A customized phage library that consists of mainly cancer-related epitopes could be a way for the improvement of its sensitivity and specificity of specific viruses."
Oncology Times: What limitations of the current study remain to be addressed to clarify the true clinical value of VES analysis in HCC risk assessment?
Wang: "The current HCC-VES has only be retrospectively validated in a small cohort of 173 at-risk patients among whom 44 developed HCC from a prospective longitudinal study. A prospective randomized study with a large at-risk population for HCC is warranted in order to evaluate a utility of VES in HCC surveillance as whether it provides survival benefit to HCC patients."
Oncology Times: Looking further to the future, what are some potential commercial applications and competitive advantages of this technology?
Wang: "This is a unique paradigm shift approach in liver cancer diagnosis. The identified VES can identify HCC prior to a clinical diagnosis from at-risk individuals. As a majority of HCC are diagnosed from at-risk individuals with chronic liver diseases due to viral hepatitis B and C, alcohol, chemical exposures and obesity, only a fraction of these individuals eventually develop HCC. Thus, it is important to determine who has the highest risk and most likelihood candidates for the development of HCC. This technology may be helpful in identifying them and recommend them for a more aggressive screening. These 'high-risk' individuals could also benefit from cancer prevention studies. The technology could serve as an effective platform for HCC surveillance."
Dibash Kumar Das is a contributing writer.