James Downing, MD, Scientific Director, St. Jude Children’s Research Hospital

Moving Forward: Phase II of the Pediatric Cancer Genome Project

As the final year of the Pediatric Cancer Genome Project (PCGP) winds down, and nearly all of the project’s goals have been either met or surpassed, one question remains –– What’s next?

Scientific Director James R. Downing, MD, who spearheaded the project, answered that question and more in an interview about St. Jude’s plans to build upon the knowledge, expertise, and infrastructure developed during the PCGP and to further investigate the genetic bases of pediatric cancer.

Will there be a phase II of the PCGP?

Work on phase II has already started. The project is a 2-year, $30 million endeavor that consists of two primary aspects: a Discovery project and a Clinical Genomics project. The Genome Institute at Washington University will continue to partner with us on the Discovery project; however, the Clinical Genomics study will be done solely at St. Jude.

Will the Discovery project simply continue the work from PCGP phase I by increasing the number of tumor samples or types of cancer analyzed via whole-genome sequencing?

During phase I, our work focused on identifying mutations that affect annotated genes or regulatory RNAs. We did not in any systematic way try to identify mutations that affect the regulatory regions that control gene expression. It is clear from both published data and the data generated through the PCGP that mutations affecting transcriptional regulatory regions play an important role in cancer.

In the second phase of the project, we will focus on defining mutations that occur in gene regulatory regions that directly contribute to tumor formation, characterize the state of the cancer cell’s epigenome, and elucidate how it differs from that seen in the normal cells from which the individual subtypes of cancer arise.

Epigenetic changes do not directly alter the DNA code but instead alter the three-dimensional structure of the DNA within a cell. These alterations occur through chemical modification of individual nucleotides or through changes in the proteins involved in packaging DNA into the nucleus. Epigenetic changes directly influence which genes are expressed within a cell and at what level.

Whereas the initial phase of the PCGP was designed to catalog the most commonly mutated genes found within the major subtypes of pediatric cancer, the phase II Discovery project will look much deeper at a smaller subset of pediatric cancers. Specifically, we plan to analyze approximately 300 pediatric cancers composed of an equal number of leukemias, brain cancers, and solid malignancies.

Each case will be analyzed for mutations by using a combination of whole-genome sequencing (WGS), whole-exome sequencing (WES), and RNA sequencing (RNA-seq). By integrating the data from these three different platforms, we will gain a much clearer understanding of the somatic mutations that exist within the specific cancer.

We will also perform whole-genome bisulfite sequencing (WGBS) of the tumor to define its DNA-methylation pattern. The latter information will be compared with a similar analysis performed on the normal cell of origin for the specific cancer subtype.

In addition, we will engraft a subset of the pediatric cancers in mice (so-called xenografts) and use those tumor models to explore the epigenetic changes that occur in DNA-binding proteins (i.e., histones and transcriptional regulatory proteins).

The integrated analysis across each of these data types will provide an unparalleled view of the genetic and epigenetic landscape of pediatric cancer. This effort will help to establish the rules for defining driver mutations in gene regulatory regions and how these mutations affect the epigenetic landscape of a cancer cell and its gene expression signature.

What are the key challenges of the Discovery project?

St. Jude is at the cutting edge of genomics research in terms of what can be done. Developing the technologies to obtain data on a large scale and the computational methods to accurately interpret that massive quantity of data are the biggest challenges.

Once a recurrent mutation in a noncoding region is identified, we must determine if the mutation affects gene expression and is thus of functional significance, or alternatively, if it simply represents a passenger mutation and should be ignored. This will require not only the integrated analysis that is described above but also direct functional experiments on the identified gene regulatory mutations by using experimental model systems.

Similarly, defining the pattern of DNA methylation at single base–pair resolution will be challenging in terms of not only generating the data but also interpreting its meaning. Key to the latter will be developing the computational methods to integrate the information gained from all of the different kinds of sequencing methods and from the tumor xenografts. Despite the challenges that exist, we feel that St. Jude is optimally positioned to tackle these questions.

What is the goal of the Clinical Genomics project?

The goal of the Clinical Genomics study is to define the optimal way to incorporate next-generation sequencing analysis into the diagnostic work-up of our patients and to determine the clinical value it brings to our efforts to cure every child we treat. Through the efforts of the PCGP, we have learned a tremendous amount about pediatric cancer. We are now poised to take this from a discovery effort into the clinic.

Although one might think this should be straightforward, it is an incredibly complicated process that ranges from sample processing and sequencing in a clinically certified laboratory, to the accurate analysis of the data generated using computational algorithms and associated databases that meet clinical standards, to the development of methods to communicate the results to the physicians caring for the patients and to the patients and their families. At every step of the process, new standards of performance will need to be established.

The Clinical Genomics project will be pursued in two stages. We will begin with a pilot study in which we will use WGS, WES, and RNA-seq to analyze 80 pediatric cancers. Approximately half of these samples have already been sequenced as part of the first phase of the PCGP, whereas the other half have not. Importantly, all of the pilot samples have gone through molecular diagnostic analyses for the molecular lesions that are currently used in routine diagnostics.

Through the pilot project, we will establish all of the methodological and analytical processes required to use next-generation sequencing in the clinical setting. The data generated will establish the accuracy of the approach for detecting the molecular genetic lesions we are currently using in the clinical setting. This will include lesions such as BCR-ABL1, TEL-AML1, E2A-PBX1, and MLL rearrangements for leukemia risk stratification and similar lists of genetic lesions associated with brain tumors and pediatric solid malignancies.

The approach will also deliver the genetic status of other cancer genes in the tumor sample and within the patient’s normal DNA, which will be analyzed in parallel. This effort is well under way.

Once the pilot project is completed, we will then move into a prospective trial in early 2014, where we plan on performing WGS, WES, and RNA-seq on all new patients with cancer who enter St. Jude during the calendar year. Although St. Jude admits approximately 450 patients each year, about 350 patients are expected to have sufficient primary tumor tissue for enrollment.

The study will require the development of a protocol, an informed consent form, and the participation of a multidisciplinary clinical team, including a clinical geneticist and genetic counselors. The data generated will be added to the data collected through the clinical protocols on which the patients are enrolled.

Although these sequence data will not be used directly for clinical decision making, they will provide a wealth of new information that should accelerate progress in advancing cures for pediatric cancer. Ultimately, this study will allow us to define the correct way of incorporating next-generation sequencing analysis into the routine work-up of pediatric patients with cancer.

What are the major challenges of the Clinical Genomics project?

The logistics of establishing the infrastructure for performing and communicating the results of next-generation sequencing analysis are daunting. We have already set up a clinical next-generation sequencing laboratory and a $4 million high-performance computer that will be used exclusively for this effort.

The laboratory is staffed by state-licensed clinical laboratory technologists, and the computational pipeline for sample analyses has been developed to meet the requirements for certification by CLIA (Clinical Laboratory Improvement Amendments). The laboratory houses five HighSeq DNA-sequencing instruments and one MiSeq DNA sequencer that will allow staff to analyze up to 20 cases per week. All aspects of the process have been tested through the pilot project, and we are on track for CLIA certification of the laboratory and computational pipeline toward the end of 2013.

Although generating and interpreting the data will be hard, it will be just as challenging to define how best to communicate the result to physicians, patients, and parents. Which results should be communicated and when may differ based on the patient’s age, sex, specific cancer diagnosis, and family history.

These and other issues will need to be studied within the context of a clinical research protocol. We believe that St. Jude is optimally positioned to address these questions so that ultimately every child with cancer in the world can benefit from the application of next-generation sequencing.

Should we expect PCGP Phase III in 2015?

The application of comprehensive DNA-sequencing technologies will become an indispensable part of both biomedical research and clinical medicine. The technology and computational methods will continue to improve.

Through efforts like PCGP and PCGP Phase II, St. Jude will remain at the forefront of the application of genomics to the advancement of the treatment of pediatric catastrophic diseases. We anticipate that over the next few years DNA sequencing projects in pediatric infectious diseases and nonmalignant hematological diseases will be initiated to complement the ongoing pediatric cancer genome effort.