Winning the War on Cancer with Big Data
By Lloyd Marino
What’s the first thing that comes to mind when you hear those words? Do you start thinking about enterprise and leveraging and analyzing information to increase the bottom line? Probably. Of course, we’re still only scratching the Big Data surface. In reality, the possibilities for Big Data go much further. From talking appliances to Project Narwhal—a new way of collecting and organizing political data—we’re also using the Big Data explosion to expand humanity’s horizons–from exploring the oceans’ deepest recesses to learning more about global warming.
Today, Big Data is turning up on the frontlines in the fight against cancer. The war against cancer is the search for medicine’s Holy Grail. Despite a number of promising advances, particularly on the prevention and treatment fronts, cancer remains a brutal illness, directly or indirectly affecting a majority of the world’s population. Few people go through life without being touched by cancer in some fashion. This year, eight million people around the globe will die, and another 14 million will receive a diagnosis of cancer. In the U.S. alone, more than 1.6 million people will be newly diagnosed and almost 600,000 will perish—grim statistics, for sure.
Given the urgency of the situation, it’s no great shock that leading biotech companies are leveraging Big Data in a number of ways to improve cancer care, identify those most at risk, and of course, produce new and better cures.
At the forefront is the American Society for Clinical Oncology’s CancerLinQ initiative, a massive undertaking that collects data from every U.S. cancer patient, making it readily available for scientific study. This information should prove especially helpful to treatment providers, giving them access to the latest real time information on how patients have fared on a particular treatment plan. Armed with this information, doctors will be able to tailor more individualized treatments, providing the best chance of a cure.
On a separate but related front, two well-funded oncology startups, Flatiron Health, a Google-funded cloud-based platform that amasses data from cancer treatment centers around the world, and Guardant Health, manufacturer of a blood-based cancer screening product that can be parlayed to more effectively stockpile patient data, have announced a strategic collaboration to build a cloud-based database filled with genomic (genetic) information, clinical trials and drug treatments, leading perhaps to the development of new, targeted cancer therapies. Currently, the only way to access genomic information about a nascent tumor is via multiple biopsies, which are expensive—the median price for a single procedure is $3700 and costs frequently escalate into five-figure territory—painful and invasive.
The hope is that the new database will give researchers and others involved in the development of cancer treatment the chance to uncover new molecular targets, streamline the recruitment and selection of new clinical trial patients, and monitor patient progress. Even patients who initially respond well to treatment relapse if their cancer evolves and mutates, which it does often, or the drugs they’re taking stop working, something that also happens, sadly. Cancer is a double-edged sword, in large measure, because it’s both a parasite (a living organism that depends on a host for survival), and an independent living organism that has a nasty of habit of ignoring its hosts’ commands.
What both Guardant and Flatiron needed was a platform to combine their unique contributions to the war against cancer. According to Fortune Magazine, only a small percentage of available cancer patient treatment data is collected systematically. And that collection largely takes place during randomized clinical trials, which cover just a fraction—4 percent to be exact—of adult cancer patients. But how can we help the remaining 96 percent? Guardant’s cutting edge blood test actually allows doctors to access a patient’s complete genomic information. Fortunately, tumors are big talkers—so to speak—and eagerly share a wealth of information about themselves. As tumors grow, they shed dead cells—no different than we do when we slough off dead skin—composed of unique genetic tissue, into their hosts’ blood stream. Using just a single tube of blood, Guardant can paint a complete genomic picture of a patient’s cancer and how it’s progressing, making it easier to target the disease. The new cloud-based database will enable Guardant to determine the results of cancer drug clinical trials in a more time-and cost-efficient manner. Why this latest advance is so important is that it takes almost nine years for a new drug just to make it to an aforementioned clinical trial—meaning where it’s tested on human beings—and 40 percent of clinical trials tank because they simply can’t find enough volunteers to participate. Using the new database, Guardant can take hundreds of its revolutionary blood tests, send the results anywhere in the world, and get the precise molecular information needed to match specific patients with the appropriate clinical trials in record time. Guardant and Flatiron will be aggregating the results from all of these clinical trials to gain insights about which targeted therapies work best for which patients.
As amazing as it sounds, thousands of patients are prescribed FDA-approved medications that may not even be warranted for their type of cancer. Sometimes these therapies work, but serendipity isn’t the best cancer fighting strategy. A dearth of clinical trial data means that physicians and patients are blindly experimenting with very expensive and even potentially useless drugs.
How is all of this playing out in the real world? To date, 14 separate North American cancer institutes have announced the use of big data analytics to match cancer patients with the treatments most likely to benefit them. Along with recommending the best drug to treat a particular cancer, analytics can also be used to suggest drugs that haven’t been used before. Armed with precise genomic information, as well as in-depth knowledge of how a certain treatment might interact with an individual patient’s specific cancer type, physicians can more confidently recommend which treatment will produce the best outcome. Though dozens of variables factor in cancer treatment, proper use of big data analytics will certainly speed up the process, saving countless lives along the way. What could be better than that?
Image by Sebastian Muller