At $US90 billion annually, the direct medical costs of cancer are among the highest of all conditions. However, rapid advances in cloud computing, next-gen DNA sequencing (NGS), and electronic medical records (EMR’s) are forming a powerful trifecta that could change the paradigm over the next decade.
It is feasible to assume that 20% of all new cancer patients in the US will receive genomic profiling by the year 2020. If cost curves play out as expected, it will cost $30 million annually to store all of this data. For the first time in history, physicians will soon have the ability to use genomic profiling for every cancer patient.
Why It Matters?
Since cancer is a disease rooted in genetics, physicians will be able to find utility in the ability to gather NGS data across large populations, analyze the data over time as the DNA mutates, and then compare it to a single patient to assess that person’s treatment options.
At the same time, these data can be used by drug companies and scientists to develop better drugs faster and cheaper than ever before. Taken together with the emergence of EMR’s, we think payers will likewise be able to translate this data into more efficient allocation of healthcare spend both for preventative and ongoing treatment-based applications.
What should be stored?
Today, most labs only retain basic gene variant data on file, not the entire genome. However, it is well established that most cancers involve multiple genes and future variant calling software tools are likely to show good improvement, so regulators have stated it is better to have an individual’s full sequence profile stored.
Harmonisation between databases is critical
Today, most genetic databases exist in silos with little in the way of data standardisation and interconnectivity between institutions. Regulators and governing bodies such as the Actionable Genome Consortium will work to develop standard nomenclatures for variant calling, standard operating procedures for data collection, and training for curators. All of this should help reduce discordance between databases.
Automation too soon is dangerous
Currently, bioinformatics platforms require human oversight, including manual curation. Regulators desire improved partnerships with the private sector where platforms are updated and validated for clinical use.
Isaac Ro is a life sciences research analyst in Goldman Sachs Research.