How Can We Study Genetic Heterogeneity Within Hepatocellular Carcinomas?
Researchers have developed a patient-derived cell line-based model to study how intratumor heterogeneity affects sensitivity of hepatocellular carcinoma (HCC) to different therapeutic agents, reported in the January issue of Gastroenterology.
HCCs have a large amount of genetic heterogeneity within each tumor, making them difficult to treat. Multiple regions of tumors can be analyzed for genetic mutations and drug sensitivity, using primary cancer cells derived from patients.
Qiang Gao et al investigated HCC intratumor genomic diversity and evaluated how it affected the responses of HCCs to therapeutic agents. The authors collected tissue samples from multiple (4–9) regions of 10 resected hepatitis B virus (HBV)-related HCCs, from patients who underwent curative resection before any adjuvant therapy. Gao et al cultured primary cells from the specimens, performed whole-exome sequence and copy-number analyses, and tested their responses to therapeutic agents (such as afatinib, bosutinib, cabozantinib, crizotinib, dabrafenib, dasatinib, ganetespib, linifanib, masitinib, sunitinib, and volasertib).
The authors identified a total of 3670 non-silent mutations (3192 missense, 94 splice-site variants, and 222 insertions or deletions) in the tumor samples. They observed considerable intratumor heterogeneity and branched evolution in all 10 tumors; the mean percentage of heterogeneous mutations in each tumor was 39.7%.
Gao et al identified 26 putative HCC driver mutations. Fewer than half of these mutations were classified as trunk events (which develop during early stages of tumorigenesis)—the remaining mapped to branches, indicating a need for therapeutic strategies that target tumor cell subclones.
All 10 patients had mutations in TP53, associated with HBV infection. Mutations in the promoter region of TERT were detected in 6 of the 10 tumors, previously reported to be an early event in tumorigenesis associated with aggressive cancers and poor outcomes of patients.
Gao et al found significant shifts toward C>T and C>G substitutions in branches of phylogenetic trees among samples from each tumor.
They found subregions of 4 HCCs to contain genetic alterations that could be targeted by existing pharmacologic agents (such as those in FGF19, DDR2, PDGFRA, and TOP1) and that were also sensitive to these agents. Cells from subregions that did not have these alterations were not sensitive to these drugs. However, high-throughput screening identified pharmacologic agents to which these cells were sensitive (see figure).
For example, overexpression of FGF19 correlated with sensitivity of cells to an inhibitor of FGFR 4; this observation was validated in 105 liver cancer cell lines.
Gao et al mention that, since they analyzed only small fractions of whole tumors, they are likely to be underestimating the true extent of intratumor genetic diversity and subclone compositions within HCCs.
In an editorial that accompanies the article, Patricia Munoz-Garrido and Jesper B. Andersen write that the findings remind us of the importance of considering a tumor’s branched evolution when trying to understand drug response.
They say that Gao et al have developed a comprehensive patient-derived cancer cell-based model that enables direct testing of intratumor heterogeneity and response to therapeutic agents. However, although the patient-derived primary tumor cells remain genetically stable through 30 passages and resemble the primary tumor, it is not clear how they are affected by loss of the microenvironment.