Analysis of Intratumor Heterogeneity in Liver Cancer Using Pharmacogenomics: A Study Based on a Large-Scale Organoid Biobank

Academic Background

Primary liver cancer (PLC) is the third leading cause of cancer-related deaths worldwide, mainly comprising hepatocellular carcinoma (HCC), intrahepatic cholangiocarcinoma (ICC), and a mixed form called combined hepatocellular-cholangiocarcinoma (CHC). The presence of heterogeneity significantly challenges precision treatment for primary liver cancer. Previous studies indicate that the genomic heterogeneity in different regions of liver cancer largely affects drug sensitivity, leading to treatment failure.

Patient-Derived Organoids (PDOs) have proven to be powerful tools for simulating tumor heterogeneity and studying drug sensitivity. However, previous studies involving liver cancer organoids are limited by sample size and lack of multi-regional samples. Therefore, this study aims to establish a PLC biobank comprising 399 organoids derived from 144 liver cancer patients, systematically dissect the genomic and phenotypic heterogeneity of PLC, screen clinically relevant drugs, and elucidate mechanisms underlying drug resistance, providing new perspectives for the precision treatment of liver cancer.

Source and Researchers

This study was conducted by Hui Yang, Jinghui Cheng, and their collaborators from the Translational Cancer Research Center of Peking University First Hospital. The paper was published on April 8, 2024, in the journal Cancer Cell, by Elsevier Inc. Corresponding authors include Jiangong Zhang from the Affiliated Cancer Hospital of Zhengzhou University and Henan Cancer Hospital, Jianmin Wu from Peking University First Hospital, and Ning Zhang. The paper is available in open access format.

Detailed Research Workflow

a) Study Workflow

1. Multi-Regional Tissue Sampling and Organoid Culturing: Multiple regions of tissue samples were collected from resected specimens of 144 liver cancer patients, totaling 522 primary tumor regions, 6 liver metastasis regions, and 30 adjacent liver tissue regions. Ultimately, 399 tumor organoids (total success rate of 75.6%) and 12 normal organoids were successfully established.

2. Genomic and Transcriptomic Sequencing: Whole-Exome Sequencing (WES) and RNA sequencing (RNA-Seq) were conducted on 99 pairs of organoids and their originating tumor tissues (from 36 patients). Additionally, RNA-Seq and drug screening were performed on 255 organoids to develop biomarkers for predicting drug response.

3. Drug Screening and Comparison: Seven liver cancer-related clinical drugs were screened on 376 organoids, calculating their IC50 (half-maximal inhibitory concentration) and Normalized Area Under the Curve (AUC).

4. Feature Analysis and Mechanism of Drug Resistance: Machine learning methods were used to analyze drug responses, identifying and validating multi-gene expression biomarkers associated with drug sensitivity, including c-Jun as a major mediator of resistance to lenvatinib.

b) Research Outcomes

1. The Organoid Biobank Recapitulates the Histological, Genomic, and Transcriptomic Features of Liver Cancer: Based on histological marker analysis, the organoids accurately exhibited hallmark features of HCC (Heppar1/AFP markers) and ICC (Krt19/EpCAM markers). Whole-exome sequencing showed high consistency between somatic mutations in organoids and primary tumors (median consistency of 87.5%), retaining most clonal and subclonal mutations of liver cancer-related genes. RNA-Seq analysis also demonstrated a high correlation at the transcriptomic level between organoids and tumor tissues.

2. Multi-Regional Organoid Analysis Reveals Genomic Heterogeneity and Its Function in Liver Cancer: Organoids from 32 patients exhibited significant genomic heterogeneity. Phylogenetic trees constructed using maximum parsimony algorithms pointed out typical truncal mutation events in PLC, such as TP53, RB1, AXIN1, and CCND1. Additionally, higher genomic heterogeneity was associated with poorer patient prognosis and drug sensitivity.

3. Drug Screening Predicts Patient Responses and Unveils Intratumor Heterogeneity in Drug Sensitivity: Screening results for the seven drugs showed high consistency between patient-level drug sensitivity and clinical responses, verifying the predictive value of organoid screening. Moreover, drug screening results were validated in organoid transplantation xenograft models.

4. Development of Multi-Gene Expression Biomarkers Related to Drug Sensitivity: Using machine learning analysis, multi-gene expression biomarkers predicting responses to four anticancer drugs, including lenvatinib, sorafenib, regorafenib, and apatinib, were developed and validated with high predictive accuracy.

5. c-Jun Mediated Lenvatinib Resistance and Its Associated Signaling Pathways: c-Jun, regulated by JNK and β-catenin signaling pathways, promoted lenvatinib resistance. Knockdown of c-Jun restored sensitivity in resistant organoids. Additionally, a compound PKUF-01 combining c-Jun inhibitor veratramine and lenvatinib was synthesized, significantly enhancing treatment efficacy.

c) Research Conclusions and Significance

This study developed a large-scale liver cancer organoid biobank that accurately recapitulated the genomic and phenotypic heterogeneity of liver cancer, providing an important resource for precision treatment. The developed multi-gene expression biomarkers offer new strategies for predicting patient drug sensitivity, with potential clinical applications. Additionally, the study revealed the mechanism of c-Jun-mediated lenvatinib resistance, providing a theoretical basis for designing combination therapy regimens.

d) Research Highlights

• Developed a large-scale liver cancer organoid biobank that systematically disclosed intratumor heterogeneity in liver cancer.
• Utilized pharmacogenomics and machine learning to develop and validate multi-gene biomarkers predicting liver cancer treatment responses.
• Identified c-Jun as a key mediator of lenvatinib resistance, unveiling its mechanism through JNK and β-catenin signaling pathways.
• Synthesized a compound PKUF-01, which significantly improved lenvatinib’s therapeutic efficacy.

e) Other Valuable Information

• This study’s research on the association between drug sensitivity and genomic heterogeneity revealed drug resistance caused by liver cancer heterogeneity.
• The potential of combining immune checkpoint inhibitors (ICIs) and anti-angiogenic tyrosine kinase inhibitors (TKIs) for future combination therapies offers new directions.

Summary

This study systematically revealed the genomic and drug sensitivity heterogeneity of liver cancer, developed multi-gene biomarkers for predicting drug responses, and identified new mechanisms of lenvatinib resistance, providing valuable resources and theoretical support for the precision treatment of liver cancer. Future clinical validations are needed to promote personalized treatment strategies’ implementation.