St. Jude Cloud Survivorship Portal offers a treasure trove of data for discovery in childhood cancers

More than 80% of children with cancer survive five years or more owing to emerging advancements in the treatment of childhood cancer. However, survival rates can vary greatly depending on the type of cancer, the treatment, and other factors. For survivors, their experience with cancer can remain a factor in their lives even into adulthood. While cancer treatment can in many cases offer a cure, these same treatments can also contribute to long-term challenges for survivors including their physical, mental, emotional, and financial wellbeing.

St. Jude Children’s Research Hospital hosts two remarkable childhood cancer survivorship studies: The St. Jude Lifetime Cohort Study (St. Jude LIFE) and the Childhood Cancer Survivor Study (CCSS). Through these landmark studies, investigators have gathered a treasure trove of data that they can use to not only better understand the health challenges that survivors face but start to design better treatments and therapies for the patients of today to improve survivorship in the future.

St. Jude has taken a proactive approach to dealing with the wealth of data provided by these studies by creating a data analysis and visualization portal specifically for survivorship data as part of St. Jude Cloud. St. Jude Cloud is the largest public repository of pediatric cancer genomics data and was launched in 2018. It provides easy access to investigators both within and outside St. Jude to study and learn from data gathered from children with cancer as well as survivors.

Survivorship data portal is an unparalleled resource

Building the St. Jude Cloud Survivorship Portal was an extraordinary collaborative effort, bringing together specialists from different departments. Experts from Epidemiology & Cancer Control, Oncology, Psychology and Biobehavioral Sciences, Global Pediatric Medicine and Pediatric Medicine meticulously assembled the data. The engineers and bioinformaticians in Computational Biology designed and developed ways to visualize and host this data live on a web-based portal housed in the St. Jude Cloud.

The data is not just from St. Jude patients, but from national and institutional cohorts of cancer survivors, creating an unparalleled cohort. This data is nothing short of a goldmine, consisting of high-quality quality demographic, diagnosis, treatment, health outcome, genetic data and patient-reported outcomes. The data dictionary contains more than 1600 variables to choose from and more than 400 million genetic variants can be explored on the genome browser. 

Xin Zhou, PhD, St. Jude Department of Computational Biology, leads the ProteinPaint team. They have crafted a unique and immersive environment that transforms these massive heterogenous datasets into actionable insights. The portal offers easy access to both raw and curated data, making browsing, visualization, and analysis possible all in one place. With the combined efforts and collaboration with the St. Jude Cloud team, the platform allows quick and easy access without the hassle of lengthy wait times to facilitate and speed up research in this critical area.

“It's been a long and fruitful collaboration with the St. Jude Cloud team, senior faculty from other departments and a tremendous amount of engineering from my team that has enabled real-time visual analysis on ‘tera-bytes’ of data,” said Zhou. “We believe that the real impact of this work is all about data sharing. This data is easily accessible to not only the entire pediatric cancer survivor research community but also other research backgrounds.”

A digital ‘treasure chest’

What's inside this digital treasure chest? The portal merges large survivorship datasets into one resource. It boasts tools such as:

• Dictionary: A highly comprehensive hierarchical organization of the data variables.
• Summary plots: Summary statistics of data variables are computed on-the-fly and visualized through highly interactive and customizable bar charts, violin plots, scatter plots and more.
• Cumulative Incidence: Ability to analyze the cumulative incidence of chronic health conditions in a cohort which can be defined by other variables on the portal, such as treatment exposure, diagnosis group, demographics, etc.
• Regression Analysis: Linear, logistic, and Cox regression where users can select outcome and independent variables, customize variables, specify interactions between variables, apply cubic splines, and visualize the analysis results.
• Data Download: Download the data of individual survivors for any number of variables on the portal.
• Genome browser: A tool for exploring and analyzing the genetic data on the portal with the capability to perform genetic association analysis.
• Custom groups: Comparative analyses between groups using any of the features on the portal.

“Prior to the portal, researchers needed to download survivorship data as large data files, parse through the data manually to extract variables or statistics of interest, and then import the data into third-party software to analyze or visualize the data,” said Gavriel Matt, Department of Computational Biology, senior computational research scientist. “Now, the portal provides a convenient point-and-click interface that allows users to explore, analyze, and visualize survivorship data all within the portal environment.”

Empowering researchers to work on survivorship

As survivorship research continues in earnest at St. Jude, so too does development of new and innovative ways to look at survivorship data. At St. Jude, Kirsten Ness, PhD, Department of Epidemiology and Cancer Control, is co-leader of the Cancer Control and Survivorship Program and  heads the state-of-the-art Human Performance Lab where investigators assess the physical function of long-term childhood cancer survivors throughout their journey, from treatment to early remission. Through this work survivors also have access to interventions such as total-body vibration and in-person and virtual exercise programs.

“While an increasing number of survivors of childhood cancer is a testament to the benefit of modern medicine, these survivors experience higher than average health risks that still need to be better understood if we are to offer them better therapies and interventions to prevent or treat the long-term effects that they can experience,” said Ness. “The Survivorship Portal is really an unparalleled resource for research that by bringing all this data together in one place with easy-to-use tools for visualization and analysis will help investigators learn more about survivorship so that we can offer patients the best possible care and support.”

This collaboration between data visualization and survivorship research and care is already producing results. Predictive models, powered by data from the CCSS study, are helping identify childhood cancer survivors at low, moderate, and high risks of developing organ failure. This knowledge is instrumental in enabling early screening and interventions, potentially saving lives.

St. Jude is at the forefront of both healthcare data visualization and survivorship research.  The power of merging these different disciplines will continue to provide insights that improve the lives of childhood cancer patients and survivors at St. Jude, and around the world.

Explore the Data

• St. Jude Cloud Survivorship Portal