Cancer treatment is on the brink of a data revolution

Imagine picking three people from a group of 100 with a certain kind of cancer, based on who you think might respond well to a certain experimental medication, then forming an opinion of the remaining 97 people based on the ones you tested. 

It would be nearly impossible to have the resources to test all 100 people, of course, which is why you chose a smaller sample amount. As simple as it sounds, this is a boiled-down scenario of how actual cancer clinical trials play out.

In the real world, only about 3% of cancer patients participate in these trials, which track how patients respond to certain treatments and determine whether a drug is working or not.

Before the introduction of big data, the information from the other 97% of the patients remained largely untapped, leaving cancer research inexact. That's changing.

Getting digitized

Moving from paper to electronic records has been no small task for health care providers. As of 2013, about 78% of all health records were electronic.

And now that the information is available, organizations have started figuring out ways to make it useful to a broader group of health care providers. One such way is through the American Society of Clinical Oncology's CancerLinQ database. 

CancerLinQ launched as a prototype in 2012, focusing on breast cancer. Through the database, they were able to pool the electronic health records of breast cancer patients from across the US. Then they were able to take their treatment information and use it to duplicate clinical trial results. That prototype pulled in 170,000 patients.

That's a big move toward getting a chance to learn from the patients who are actually in treatment.

Dr. Clifford Hudis, who oversees the breast medicine practice at Memorial Sloan Kettering Cancer Center in New York, gave the example of breast cancer patients. Most candidates for trials are young, healthy, and highly likely to benefit from treatment. But in practice, the people who end up receiving the medications resulting from those trials do not match this profile. With the help of big data in the form of electronic health records and accessible databases, doctors could start to learn from more than just those patients used in clinical trials. 

"This wasn't possible before data was stored electronically," Hudis said.

CancerLinQ plans to fully launch for all types of cancer by 2016, Hudis said. And CancerLinQ isn't the only database like this out there, which he said validates how important compiling this information is. 

Diagnosing and monitoring cancer with big data

Big data isn't just increasing the potential for what oncologists can do with the information from hundreds of thousands of patients. Faster genetic sequencing technologies are also helping companies find easier ways to track cancer treatments at a genetic level.

Guardant Health, a company that makes genetic cancer tests for people with advanced or late-stage cancer, says it's using the power of data to offer a more accurate and less invasive alternative to traditional biopsy tests, which require a sample of tissue to test for cancerous cells, often requiring patients to undergo anesthesia.

Guardant's "liquid biopsy" 360 test takes blood samples from cancer patients and sequences the genetic information in that blood to figure out how tumors are responding to a certain course of treatment. As a result, a patient using the 360 test can get a liquid biopsy every few weeks, rather than every few months, the timeline for a traditional biopsy.

They're not the only ones taking on cancer with big data. Guardant recently partnered with Flatiron Health, an oncology technology company, to make Guardant's test information more accessible to health care providers. Other companies and researchers are working on their own liquid biopsy tests as well, such as RainDance Technologies and Qiagen. Software companies like Cypher Genomic are working on ways to sequence genes faster with the hope of speeding up cancer treatment and drug development. 

Eltoukhy said big data will be a way to "democratize" clinical trials, by providing a larger patient base that's eligible to enroll in these trials.

 "We can now massively digitize cancer," Eltoukhy said.

Just the beginning

Because these technologies are just getting off the ground, the success stories so far are only going to get better as doctors learn more about how to use the information that's out there.

"It's not that big data is the prime goal, it's filling a vacuum," Hudis said. "This is allowing us to address those gaps."

The most important thing to Hudis is that big data is just a piece of a larger picture of cancer treatment that can enhance how doctors provide care. The more informed the treatment plan is, the better the chances the cancer patients will survive.

"What big data represents is an additional resource," he said. "leaving it on the table untapped is a tragedy."