Molecular Signature Reveals Which Liver Cancer Patients May Benefit from a New Drug

The array in Figure A depicts the clustering of a subtype of  hepatocarcinoma based on its demethylation signature. A 133-gene list was reduced to 16 genes, which correctly predicted a survival subtype A.  Figure B shows the sensitivity and specificity of using the demethylation signature for predicting drug responders.   Figure C shows the cumulative survival risk prediction within subtype A.

The array in Figure A depicts the clustering of a subtype of hepatocarcinoma based on its demethylation signature. A 133-gene list was reduced to 16 genes, which correctly predicted a survival subtype A. Figure B shows the sensitivity and specificity of using the demethylation signature for predicting drug responders. Figure C shows the cumulative survival risk prediction within subtype A.

Only one drug currently on the market has the potential to extend survival for patients with advanced-stage liver cancer and only 30 percent of patients are eligible to receive it. As the fastest-growing type of cancer by incidence in the United States, liver cancer represents a major public health problem and there is an urgent need to develop new treatment strategies.

In recent years, scientists have focused their efforts on classifying cancer into subtypes by using genetic profiling combined with clinical findings. Such profiling can inform doctors about a patient's survival odds and treatment options. For example, the information can be used to determine whether a patient is likely to benefit from a particular drug.

Applying this approach to liver cancer, Jesper B. Andersen, Ph.D., from CCR's Laboratory of Experimental Carcinogenesis, working with Snorri S. Thorgerisson, M.D., Ph.D., and colleagues, analyzed gene expression changes that occur in liver cancer cells treated with an experimental drug called zebularine, which has shown promising results in several cancer cell lines, including bladder, prostate, and lung. The results of their work were published in a recent issue of Science Translational Medicine.

In their previous work, the research team determined that progression of liver cancer from premalignant to advanced stages is accompanied by both genetic and epigenetic changes. While genetic changes are irreversible, epigenetic changes can be challenged with medications such as zebularine. The drug reactivates silenced genes, including genes that suppress tumor growth, through a process called demethylation.

Analysis of human liver cancer cell lines representing various subtypes of the disease showed that some lines were sensitive and some were resistant to zebularine; the drug halted growth and triggered cell death in the sensitive cell lines but not in the resistant cell lines. The same trend was observed in tumors derived from the sensitive versus resistant cell lines, confirming zebularine's therapeutic potential in vivo. Notably, treatment of experimental animals with zebularine caused no adverse effects. Gene expression profiling revealed that in the sensitive tumors, zebularine activated genes that slow tumor growth; whereas, in resistant tumors, it activated genes associated with promoting tumor growth.

Analysis of human tumor samples identified a set of genes critical for predicting recurrence of disease and length of patient survival. Patients with the highest level of gene silencing (methylation) also had the poorest prognosis. Based on these findings, the researchers developed two zebularine response signatures: a 16-gene genomic response signature and a 133-gene epigenomic response signature. The signatures can be used to predict which liver cancer patients are likely to benefit from the zebularine therapy and how fast their disease might come back. Future studies will attempt to elucidate, on the molecular level, why some tumors are resistant and some are sensitive to epigenomic therapy.

Summary Posted: 10/2010

Reference

Sci Transl Med. 2010 Oct 20;2(54):54ra77