In just a few years, cloud technology has evolved from merely a concept to an omnipresent 21st century essential. Though the cloud is very much here on earth—essentially a network of computers that provide services through wired or wireless internet connections—its migration from board room buzzword to practical tool is now making a difference in healthcare. And it is offering new opportunities to expand diagnostic testing services, according to analysts who evaluate technology trends.
A recent report from the global market research firm IDC predicts that 80% of healthcare data worldwide will pass through the cloud at some point in its lifecycle by 2018 as part of what the firm calls third platform technologies, which include the cloud, big data and analytics, mobile internet access, and social media tools. “Third platform technology makes these tools widely available, at low cost, and in agile, responsive computing environments,” according to the report, IDC FutureScape: Worldwide Healthcare 2015 Predictions.
“The cloud has made supercomputing capabilities available to any laboratory in the world, at low cost, with global multi-site distribution capability,” added Judy Hanover, research director for provider IT transformation at IDC.
Hanover and her colleague, Alan Louie, PhD, research director for clinical development technologies and strategies at IDC, believe that the biggest impact on clinical labs will be increased utilization of big data and analytics for next-generation sequencing (NGS) and its application to personalized medicine. For example, researchers are already working on databases of treatments associated with specific genomic variants that, increasingly, the cloud will make more accessible and affordable.
Building on the cloud, mobile technology will continue to have a two-fold impact. Information from anywhere or even multiple locations can be delivered in real time to clinicians’—and patients’—mobile devices.
Louie noted the cloud also will lead to improved collaboration, making it possible for specialty labs to integrate test results and other data more easily with other labs. “When a traditional clinical lab obtains a blood sample from a patient, the lab will have new tests available through the greater lab testing ecosystem that otherwise would not be available. And this is happening now,” he said.
Supercomputers…for Super Results?
A large multidisciplinary team of researchers at the University of Minnesota’s Molecular Diagnostic Laboratory and the Supercomputing Institute for Advanced Computational Research have been working together since 2010 to design, validate, and implement cost-effective NGS data analysis and assays. Kevin Silverstein, PhD, is a lead scientist for the research informatics support systems at the institute. He explained that one of the major beneficiaries of cloud computing will be smaller diagnostic labs, as the cloud obviates the need to purchase, update, and operate costly computers. Economical access to computing power via the cloud also enables multiple samples to be sequenced simultaneously and rapidly. Most importantly for researchers, data and tests can be reliably replicated, regardless of infrastructure changes made to non-clinically dedicated computers.
“One major challenge to implement NGS-based technology in clinical labs has been the need to develop a scalable and robust bioinformatics infrastructure to effectively handle the volume of data produced,” Silverstein said. “We have developed and implemented a bioinformatics processing pipeline for the molecular diagnostic labs to help manage per-sample analytical costs, and we have developed software to control storage costs. The lab staff is frequently developing, certifying, and validating new NGS tests for the molecular diagnostic lab. The result has been a remarkable increase in the number of tests ordered here locally.”
Bharat Thyagarajan, PhD, assistant professor in the department of laboratory medicine and pathology and director of the molecular diagnostic lab at the University of Minnesota, is delighted with these developments. In 2010, the lab’s total testing menu was six gene sequences. Two years ago, this increased to 560 genes. Today the lab is now offering sequencing for 4,800 genes.
Relying on cloud computing, Thyagarajan’s lab has performed extensive validation of targeted NGS for inherited disorders, and has developed clinical test panels for 568 genes that are associated with more than 100 distinct inherited disorders.
Supercharging New Insights Into Cancer
The power and agility afforded by cloud-based computing is leading to unique collaborations and, in some cases, a new understanding of cancer. For example, Memorial Sloan Kettering Cancer Center (MSKCC) and Quest Diagnostics in June 2014 announced a joint collaboration to identify and annotate gene mutations associated with solid cancer tumors. MSKCC is providing contextual information, including potential treatment options, about individual mutations identified as part of Quest’s OncoVantage, an independently validated, laboratory-developed test for molecular characterization of solid tumors.
In addition to its collaboration with Quest, MSKCC has implemented another cloud-related innovation, cBioPortal for Cancer Genomics, a resource for the visualization and analysis of large-scale genomics data sets. Nicholas Schultz, PhD, who is leading cBioPortal’s development, explained that it takes public domain large scale data sets—sequencing profiles of thousands of cancer patients—and makes the data accessible, viewable, and interpretable for biologists throughout the world. The tool is being used by about thousands of researchers every week.
The cBioPortal has become the main tool used to visualize data on research projects at MSKCC and also for more than 5,000 patients annually who have their tumor genomes sequenced. “The ability for us to apply this tool to one patient at a time enables us to identify tumor mutations efficiently and economically,” said Schultz, who also is an associate attending computational oncologist in the department of epidemiology and biostatistics, and head of knowledge systems in the Center for Molecular Oncology at MSKCC. “We can determine which mutations we have previously seen, the frequency at which we see them, and the context in which we see them. We can identify which ones we know are actionable and have proven drug sensitivities. A lung cancer, for example, may have 100 mutations, but fewer than five are possibly oncogenic. Our goal is to identify which ones.”
MSKCC’s collaboration with Quest is accelerating the annotation process for known cancer-causing variants, according to Schultz. Fifty MSKCC scientific curators are currently annotating 400 cancer genes in detail, from the basic function of the gene background to prevalence of differences in disease types to prognostic implications of mutations in those genes. The biological and biochemical effects are annotated, and if known, drug sensitivities for Food and Drug Administration–approved drugs (currently involving about 10 genes) are noted. Data on the genes’ sensitivities to drugs in pre-clinical and clinical evaluation also are added.
According to IDC’s Louie, more clinical laboratory professionals should be taking a hard look at what the cloud means for their own information technology needs. “It is important to first take an introspective look at the IT infrastructure supporting the clinical lab overall today,” said Louie. “It will also be wise to work with well-established companies which have significant experience in this area to avoid hurdles that have already been overcome in other healthcare—and other industry—efforts. The third platform represents a major paradigm shift that will likely be a foundation for growth for the foreseeable future.”