MedicalResearch.com Interview with:
Olivier Elemento, PhD
Head, Laboratory of Cancer Systems Biology
Department of Physiology and Biophysics
Institute for Computational Biomedicine
Weill Cornell Medical College New York, NY, 10021
Medical Research: What is the background for this study? What are the main findings?
Dr. Elemento: In many cancer patients, initial treatment with chemotherapy or targeted therapy shrinks the tumor or makes it disappear; however the tumor eventually comes back in a form that is frequently resistant to treatment. This process is called relapse. In diffuse large B-cell lymphomas (an aggressive type of blood cancer), approximately 40% of patients eventually relapse. How relapse occurs and how these tumors adapt and become resistant to treatment is not well understood. Why 60% of patients do not relapse and are essentially cured of their disease while 40% relapse is not known. Many think the relapse process involves tumor cells acquiring DNA alterations that make them resistant to therapy. We have indeed previously identified such DNA mutations in diffuse large B-cell lymphomas (http://genomebiology.com/2014/15/8/432/abstract).
However it became apparent to us and others that DNA mutations do not explain fully why these tumors become treatment-resistant and come back, sometimes years after initial diagnosis and therapy. We therefore turned to the epigenome to look for possible reasons. Specifically, we studied chemical modifications of DNA called DNA methylation. DNA methylation is transmitted faithfully from one cancer cell division to the next but can also be modified by specific enzymes. DNA methylation is thought to impact the way genes are expressed in tumors by modulating accessibility of DNA to proteins that regulate gene expression. DNA methylation is known to be altered in tumors. Could DNA methylation alterations also be at least partially responsible for relapse and resistance to treatment ?
To address this question, we used a high-throughput DNA methylation profiling method to capture the DNA methylation landscape genome-wide. That is, we queried DNA methylation status at millions of locations in the tumor genome. We profiled B cell lymphoma biopsies from patients treated at Cornell and Torino. Since we were interested in how tumors change upon treatment, we profiled the initial tumor biopsy obtained at time of diagnosis (pre-treatment) then profiled the biopsy obtained at time of relapse in the same patients. The profiling produces enormous amounts of epigenomic data and we therefore had to use customized Big Data analytical algorithms together with supercomputers to interpret the methylation patterns.
Our main findings are as follows:
1) We found extensive changes in DNA methylation between diagnosis and relapse – many more epigenomic changes than DNA changes. The changes are partially random – every patient’s tumor changed in different ways. However, we observed many convergent changes. That is, near some genes, all or many patients undergo the same changes. Some of the genes are known to impact tumor biology, for example genes in the TGFbeta pathway. We also found significant epigenomic changes at genetic switches – the regions in the genome that control which genes are expressed. We think that DNA methylation changes occurring at relapse interfere with the function of these switches, and perturb expression of genes controlled by these switches.
2) Perhaps most interestingly, we found that the cell-to-cell methylation heterogeneity within tumors at time of diagnosis is highly predictive of relapse. Tumors with elevated cell-to-cell methylation heterogeneity, that is, tumors where every cell has a different methylation landscape, are more likely to relapse. The higher cell-to-cell methylation heterogeneity is, the faster tumors relapse.
These are exciting findings that for the first time indicate a major role of the epigenome in supporting the evolution and adaptation of tumors in response to treatment.
Medical Research: What should clinicians and patients take away from your report?
1) It is important to understand how tumors change in time, especially after treatment. This applies to all tumors, not only B cell lymphomas. A tumor found at relapse is not the same as the tumor found at initial diagnosis. Relapse tumors have evolved at the genomic level and even more so at the epigenomic level as our study shows. We cannot treat a relapse tumor as if it was the same tumor as the diagnosis one. The tumor changed – we need to understand what changed in order to tailor the treatment to the new tumor. This is important because for many patients, no biopsy is performed at time of relapse. This has to change.
2) The epigenome has potential clinical relevance. In our study we found that intra-tumor methylation heterogeneity early at diagnosis can predict which patients will eventually relapse and how fast they will relapse. The precise predictive power of the epigenome needs to be studied further but one may envision that eventually patients with elevated intra-tumor methylation heterogeneity will be monitored more aggressively for signs of relapse and/or given more aggressive treatment regimens since they may be at higher risk of relapse.
Medical Research: What recommendations do you have for future research as a result of this study?
Dr. Elemento: We need to better understand the role that the epigenome plays in tumor progression. What genes are affected and how? How does the tumor genome cooperate with the tumor epigenome ? Once we understand how the epigenome contributes to tumor progression, we may be able to find new ways to block or slow down such progression, maybe by modifying the epigenome. There are already FDA-approved drugs that can interfere with DNA methylation and other epigenetic mechanisms. Maybe such drugs will be used one day to limit the ability of tumors to evolve and adapt.
At a time where precision medicine programs are put in place in hospitals across the country, we think it will be important to not only sequence the genome of patients but also characterize their epigenome. The epigenome may help understand the blueprint and alterations that led to disease (not only cancer) and provide a rich source of clinically actionable information.
Olivier Elemento et al. Epigenomic evolution in diffuse large B-cell lymphomas. Nature Communications, April 2015 DOI: 10.1038/ncomms7921
MedicalResearch.com Interview with: Olivier Elemento, PhD (2015). Epigenetic Changes May Account For How Some Lymphoma Tumors Change Over Time