Molecular information: Making data meaningful

The healthcare industry is facing a Big Data challenge

Key global medical conferences such as the European Cancer Congress (ECC) and the American Society of Clinical Oncology (ASCO) Annual Meeting provide the medical community with the opportunity to present and discuss the very latest progress in oncology testing and treatment. Most of the data presented are generated from patients participating in clinical trials. What is missing, however, is how these data can be connected to all of the other health information available today to give doctors and researchers a comprehensive view of a patient’s disease. Largely unnoticed, there is a quiet revolution taking place within healthcare that could see a paradigm shift in our approach to cancer care.

We are currently seeing an explosion of healthcare information provided by modern technologies that is generating a wealth of new data and insights in the cancer field. To illustrate this point, consider genomic sequencing: back in 2003 it took over two years and cost three billion Swiss Francs to sequence the human genome. In just over a decade this can now be achieved in a few hours for only a few thousand Francs

Broader availability of sequencing will lead to greater collection of genomics data about individuals. These data, combined with other information obtained from electronic medical records and other forms of diagnostic testing, will contain valuable insights about individuals which will need to be stored, transferred and analysed.

The power of sequencing can transform the way we approach healthcare and expand personalised medicine.
Roland Diggelmann, COO Roche Diagnostics.

We are entering the era of Big Data

  • Cancer is no longer seen as just one disease, but a collection of hundreds of diseases with different characteristics and genetic make-up.
  • The number of disease-driving mutations that can be detected and potentially treated has greatly increased – e.g. in non-small cell lung cancer from two in 2004 to over ten today – and is expected to grow even further (see figure below).
  • Today large databases, hospitals and patient monitoring generate an enormous amount of data on thousands of patients.
  • Global data will soon reach the zettabyte scale (that is about the same as 36 million years of HD video).
  • 80% of medical data is unstructured yet clinically relevant, i.e. there are huge potential gains to be made by looking at it all as a whole
  • A decade ago, mutations were important in only around 1/3 of non-small cell lung cancer cases
    A decade ago, mutations were important in only around 1/3 of non-small cell lung cancer cases
  • Today, there are many known mutations that can drive the development of cancer...
    Today, there are many known mutations that can drive the development of cancer...
  • ...which can all overlap, significantly increasing the complexity of the disease
    ...which can all overlap, significantly increasing the complexity of the disease

From the sequencing of the human genome to the output from sophisticated diagnostic tools, to the patient and treatment databases behind clinical trials and registries, as well as the large and ever-growing number of healthcare-related smartphone apps used worldwide, there are an enormous number of data sources now available. These are becoming so vast and complex that novel IT tools and technologies are required to capture, manage and analyse them. In addition, there are many open questions related to the standardisation and compatibility of data from different sources, as well as many associated data privacy concerns. Therefore, a major challenge for the healthcare industry is how to work best with these data in order to gain new insights; spotting correlations and trends which could lead to better patient outcomes.

Molecular Information stands to revolutionise how we look at cancer

Data on genes, proteins and other factors are referred to collectively as Molecular Information, and changes and interactions between these data can provide insights about individuals, their diseases and which treatments could best help them.

New technological possibilities to collect and analyse Molecular Information could broaden the options available for diagnosing and treating diseases, by going beyond traditional companion diagnostics. For example, oncologists may be able to characterise the complete genetic profile of a patient’s disease using minimal tumour tissue (obtained from a biopsy), which can be difficult to collect in sufficient amounts for comprehensive testing. New technologies may also allow doctors to monitor how a patient’s cancer changes over time. This would allow doctors to select the very best treatment plan for a specific patient, thus representing personalised healthcare in its purest form.

Furthermore, anonymised patient data could be grouped together and used to identify new potential targets for therapy, allowing researchers to focus their efforts on developing new treatment approaches in areas with the greatest unmet need.

Molecular Information is about taking standardised information across different tumour types and patients, putting it into a database, and then being able to extract meaningful information for patient care on the one hand, and generating R&D hypotheses on the other. It will help us make better decisions and change the standard of care.
Daniel O’Day, COO Roche Pharma.

Roche is leading the way in making best use of Molecular Information

Roche, as a pioneer in personalised healthcare, recognises Molecular Information as the logical evolution in this area, and is already taking steps to maximise the potential benefits from this approach. For example, in 2014–2015 we acquired or formed partnerships with a number of companies with novel approaches to healthcare information. We are now working to integrate these companies, such as Foundation Medicine (which specialises in genomic profile tests) and Bina (a bioinformatics company), into our work to be able to benefit millions of patients around the world.

With proven expertise in both pharmaceuticals and diagnostics and strong relationships with key industry partners, Roche is strongly positioned to master the complexity of this new data-driven frontier in healthcare. We will continue to embrace ground-breaking technologies not only to drive the data revolution, but ultimately to take personalised healthcare in oncology to the next level, thereby driving a step change in patient outcomes.

There is no doubt that in five or ten years, the treatment of cancer will look very different. The ability to bring pharmaceuticals and diagnostics together in a disease area that is so complex puts us as a company in a privileged position to make a real contribution for patients.
Severin Schwan, CEO Roche