Two Chinese-American Scientists Tackle the Challenges of Early Cancer Detection with Innovative Liquid Biopsy Solutions

In 2016, the prestigious oncology journal CA: A Cancer Journal for Clinicians published a report on estimated cancer statistics in China for 2015. The data revealed that there were 4.292 million new cancer cases and 2.814 million cancer-related deaths in China in 2015. According to cancer statistics released by the World Health Organization, there are approximately 14 million new cancer cases and 8.8 million cancer-related deaths worldwide each year. Based on these figures, China’s cancer-related deaths account for more than one-quarter of the global total, underscoring its status as a country with a high cancer burden.

From 2012 to 2016, the number of cancer survivors in the United States increased by 1 million. Nevertheless, in 2016, there were still approximately 1.685 million new cancer cases and 596,000 cancer-related deaths. Despite variations across countries and regions, and although modern medicine and scientific advancements have significantly extended human life expectancy, cancer remains a global public health threat.

For certain cancers, the difference in patient survival rates between early and late diagnosis is substantial. Taking liver cancer as an example, the five-year survival rate is as high as 70% if diagnosed at an early stage, whereas it drops to less than 5% if diagnosed at a late stage. Therefore, early cancer screening can be regarded as a critical first step in the fight against cancer.

Liquid biopsy-based technologies for early cancer screening first emerged in the United States. Compared with traditional tissue biopsy, they are significantly less invasive and have far fewer side effects—offering new hope for early cancer detection.

“However, there are currently very few liquid biopsy companies domestically and internationally that focus on early cancer screening,”Most are focused on the detection of mid- to late-stage tumors and guidance on medication.“Professor Li Bingshan from the Institute for Medicine and Public Health at Vanderbilt University told VCBeat. He then explained to reporters that, compared with patients in advanced stages, early-stage patients have very low levels of cell-free DNA derived from cancer cells in their blood, which in most cases fall below the detection limits of current methods. This poses significant challenges both technically and in terms of subsequent data analysis.”

“Moreover, early-stage tumor screening also requires large-scale prospective patient samples,” added Professor Hu Shan Yang of the Sidney Kimmel Cancer Center at Thomas Jefferson University in Philadelphia. “For instance, Grail, which recently secured financing, has been engaged in this very effort over the past two years.” Clinical trials for early cancer screening must be conducted on large-scale sample cohorts to validate the accuracy of screening technologies. Although it started earlier,Most overseas liquid biopsy companies for early cancer screening are currently in the early stages of technology development and resource accumulation., such as Grail, Freenome, and Guardant Health.

The two professors met at Baylor College of Medicine in Houston, USA, and have known each other for nearly 20 years. Both researchers focus on large-scale population data; one specializes in cancer epidemiology and biomarker development, while the other focuses on statistical genetics and next-generation sequencing algorithm development. Together, they form a perfect combination for cancer early screening research.

After earning their Ph.D. degrees at Baylor College of Medicine, the two researchers completed their postdoctoral training at the MD Anderson Cancer Center and the Center for Statistical Genetics at the University of Michigan, respectively. Approximately seven to eight years ago, they each established their own independent research laboratories. During the development of their labs, they received substantial funding from institutions such as the National Cancer Institute (NCI), the National Human Genome Research Institute (NHGRI), and the American Cancer Society (ACS).

Since establishing their respective laboratories, the two have collaborated on risk prediction and early screening for liver cancer, publishing multiple research papers. In 2012, a relative of Professor Yang Hushan was diagnosed with liver cancer before the age of 50 due to long-term chronic hepatitis B infection. Due to the late detection, this relative passed away shortly thereafter. Professor Yang recalled that this incident had a profound impact on him. Although he had been engaged in research on early screening for liver cancer, he was unable to prevent such a tragedy from befalling his own family. After an in-depth discussion with Professor Li Bingshan, they resolved to accelerate the translation of their findings into clinically applicable early-screening products, helping more families avoid similar regrets. Through their prior research, they gradually recognized that their accumulated patient resources, along with the technologies and algorithms they had developed, could be practically applied in clinical settings to screen hepatitis B patients for the onset of liver cancer.

The regret of losing loved ones has also become a driving force in clinical translational research.China has approximately 100 million patients with chronic hepatitis B, and their incidence of liver cancer is more than ten times higher than that of the general population.This type of cancer, which has a high incidence in China, is relatively rare in the United States, and there are few effective products available for early screening. The current clinical approach, which combines alpha-fetoprotein testing with imaging and liver biopsy, does not offer high accuracy and is limited by factors such as the invasive nature of biopsies and tumor heterogeneity. Therefore, they initially focused on developing non-invasive liquid biopsy technologies to predict hepatocellular carcinoma risk and enable early diagnosis among patients with hepatitis B.

Based on years of dedication to the field of early cancer screening, they believe that early-stage and late-stage cancers exhibit markedly different characteristics. This is why many detection methods developed based on the features of late-stage cancer are not applicable to early-stage cancer. They argue that the most critical task at present is to uncover the genomic characteristics of early-stage cancer and to develop multi-omics diagnostic strategies tailored to these features, leveraging integrated models to maximize diagnostic accuracy. Furthermore, since trace amounts of cancer-related genetic DNA alterations can also be detected in the blood of many healthy individuals, early cancer screening technologies must undergo validation using large-scale clinical samples, particularly those from patients with early-stage tumors or even pre-clinical diagnosis stages. Otherwise, even highly precise techniques may yield signals that cannot effectively confirm whether the individual actually has cancer.

The team’s recent research outcome is iDriver, a precision oncology decision-support tool based on integrated analysis of large-scale medical big data, published in the prestigious journal Bioinformatics. This technology leverages unsupervised machine learning algorithms and Bayesian statistical models to integrate multi-omics data from a large cohort of cancer patients across diverse types, including genomic mutations, copy number variations, gene expression, protein expression, and methylation profiles. By comparing and modeling these data against multi-omics data from extensive normal tissue samples, the tool rapidly identifies cancer driver genes in patients, thereby accurately predicting their cancer types.

In the early stages of tumors, genetic mutations are relatively scarce and are predominantly enriched in driver gene variants. iDriver effectively integrates multi-omics big data to identify a more comprehensive set of driver variants, thereby providing valuable assistance in the early diagnosis of cancer. Currently, this model is primarily applied to the mining of public big data, such as The Cancer Genome Atlas (TCGA) database, and quantifies the strength of all driver genes. On one hand, it can provide new candidate genes, laying the foundation for functional studies by researchers; on the other hand, it can integrate associations across different genes and omics layers, effectively improving the accuracy of early cancer diagnosis. Professor Li Bingshan revealed that his team is currently expanding the application scope of the model, aiming to integrate individual multi-omics data to further refine the prediction of personalized driver genes, thus playing a guiding role in early cancer screening and precision therapy.

Furthermore, the research team has further refined a liquid biopsy technique based on genome-wide chromosomal copy number aberrations in tumors, with early screening performance that has already surpassed certain existing clinical indicators. This technology builds upon the team’s research in the field of early liver cancer screening and leverages large-scale prospective samples collected over many years from both clinical diagnoses and pre-diagnostic stages. The two researchers also disclosed to VCBeat“This may be the only technology validated in large-scale prospective cohorts today.”。

“Based on existing technologies and computational models, our research team will further develop early screening products based on multi-omics data,” Professor Yang told VCBeat. “Different omics layers can provide a wealth of complementary information, and deep integration of these data can significantly improve the accuracy of detection.”Meanwhile, they also plan to further expand the application of their technological approaches to early screening for multiple other cancers with high incidence in China, ultimately achieving pan-cancer detection in asymptomatic populations on a gradual basis.Although the team’s focus in the initial stage differs from that of U.S. early cancer screening companies such as Grail, it is evident that their ultimate goals are highly aligned.

Professor Li also revealed that a key component of the multi-dimensional omics data products under development by his team is methylation-based methodology. The team’s independently developed methylated liquid biopsy technology has achieved high sensitivity, enabling stable library construction and analysis from as little as 1 ng of circulating tumor DNA (ctDNA). This demonstrates the potential for detecting ctDNA at ultra-low concentrations. In the next phase, they will further optimize the technology, integrate additional whole-genomics information and tumor-specific biomarkers from specific genomic regions, and advance corresponding statistical models to bring the product to a level suitable for clinical use.

In addition to R&D, the team has established strategic partnerships with multiple Grade 3A oncology and infectious disease hospitals in China and will soon launch large-scale clinical studies. Professor Li told VCBeat:“Clinical studies must be conducted in China; only by accumulating data that reflects the characteristics of the Chinese population can we develop products that truly serve the Chinese people.”

During the interview, both individuals repeatedly emphasized that products should align with “the characteristics of Chinese people” and “serve the Chinese population.” Many people discuss the differences between domestic and international research; being on the other side of the ocean, they may feel these differences more acutely. China’s cancer “epidemic” is more severe than that in the United States and has distinct features. Given that the U.S. has Grail, a company dedicated to early cancer screening, shouldn’t China have one as well?