Fighting cancer with IT
In the 19th century, the use of the microscope to study the pathology of disease led to the birth of modern oncology. CIOs working in the cancer field today say we're on the verge of a similar revolution.
Two hundred years ago, "they realised there were things that couldn't be seen with the naked eye that could explain how the disease progressed," says Skarulis of Memorial Sloan Kettering. "It's that type of change we're experiencing today in terms of being able to see things we couldn't see before."
Next-generation DNA sequencing is the microscope of the 21st century. More streamlined and affordable technologies for studying the DNA of cancer cells are expected to speed up the development and implementation of improved and personalised cancer treatments.
Next-generation sequencing methods are much faster than their traditional forebears, producing millions or billions of sequences at once. And the cost is plummeting. The Human Genome Project needed US$3 billion and 13 years to sequence our DNA.
Today, researchers can sequence a genome in 10 days for US$5,000. "In the next few years, you'll be able to get a full genome sequence of your DNA for US$500," says Sorena Nadaf, associate director and CIO of the Helen Diller Family Comprehensive Cancer Center at the University of California, San Francisco (UCSF). "Pretty soon insurance will cover it like another blood draw."
In the near future, "every patient that comes in the door with a cancer diagnosis is going to get clinical-grade, next-gen sequencing," says Warren Kibbe, CIO of the National Cancer Institute (NCI) and director of the NCI Center for Biomedical Informatics and Information Technology. "In fact, it won't even be called next-gen sequencing soon."
That creates incredible IT demand. "It drives huge computational needs," says Kibbe. "It's exciting, but it's challenging because the environment can change overnight." Cloud computing holds promise, particularly for far-flung and smaller institutions.
"There is a sea change in the way we operate as IT professionals," Kibbe says. "The ability to achieve true interoperability and utilise infrastructure- and platform-as-a-service in clinical research will revolutionise our ability to determine optimal treatments for new cancer patients."
When working with the scientific community, Kibbe is dealing with with evolving problems and new technologies to find solutions that weren't possible yesterday. Those research systems are as far from enterprise IT as you can get, intimately tied up in the particular research of an individual lab. But as the science matures, Kibbe must figure out how to make those one-off solutions scalable for the rest of the community.
That adds a dimension to the CIO role. "I have to deliver an awful lot of basic services like every other CIO - networks, storage, file services, security. But I have this really exciting additional mission to take into account," he says. "On one side, it's the classic CIO blocking and tackling. On the other, it's how do we interface with the scientific community to further our mission. There's opportunity to do better on both fronts."
Very big data
Jack London, director of the informatics shared resource at the Kimmel Cancer Center, says he's been in medical informatics since before it was called informatics. A version of the first mainframe he worked with could be a museum piece. He's seen technology tools evolve to keep up with the demands of cancer research and treatment. For the last few years, he has focused on personalised cancer medicine.
"If you go back to the 1950s until very recently, the way we treat cancer is to give people a set of toxic drugs that kill the rapidly growing cells, and you adjust the dose so you don't kill the patient at the same time. It's the brute-force approach," London says.
"But cancer is a genetic disease. From an IT perspective, that means not only analysing very large sets of genomic information, but also managing specimen banks and their associated clinical data and correlating that with the latest research. "You get [up] to terabytes of data in a very short time," London says.
Then you have to figure out what to do with those terabytes. The goal is clinically actionable results.
"You have to get to point that you can put it in an electronic health record and communicate that to the clinician who is treating a patient today," London says. "That's what everyone is working very hard on."
At Memorial Sloan Kettering, physicians have been archiving samples of patients' cancer cells, along with their clinical histories, for 30 years. Skarulis oversees one of the largest single-organisation data warehouses for patient care and clinical research in the country.
Sloan Kettering's secure, Web-based system contains more than a million cancer patient records and is the basis for research into the complex mechanisms that cause cancers to form or progress. Such investigations have resulted in more effective therapies and diagnostic tools for some lung, colorectal and skin cancers.
Next-generation sequencing will take that to the next level. It "can spew out data at a rate that's incomprehensible, but that data has to be stored, processed and made available to doctors," says Skarulis, whose team rolled out a system in May to do just that.
"We've been meeting with many, many people from the institution, experts in everything from computational biology to sequencing itself, working together to pull this off," she says. "I couldn't be alive at a more exciting time in cancer research, and we're helping to get that to the bedside."
Both the technology and the science are advancing rapidly. "The explosion in technology is enabling us to examine things we couldn't look at before," Skarulis says. "One is enabling the other."
"Data," says Nadaf, "is driving change."
A team science
That explosion of data is also fostering closer relationships among CIOs in the research and clinical area.
"In the last 20 years," says Kibbe of NCI, "it's become clearer and clearer that cancer research is a team science. My job is very much the facilitator of that."
The US government paved the way for increased collaboration with the Cancer Biomedical Informatics Grid (CaBIG) in 2004, a program to develop an open-source, open-access information network for cancer research.
"Before that, everyone was working in silos," says London of the Kimmel Cancer Center, "and not only working in silos, but competing for grant dollars." As a result, groups at various institutions might have been getting funding to work on similar or duplicative projects.
"They knew they needed to get them to share their data, and they seized upon the IT infrastructure as a way to do that," London says.
Ultimately, the CaBIG program was retired, but not before "it got us all working together," he says.
London and other biomedical informatics leaders in the NCI system are cooperating in various areas to better support the scientists and clinicians at their individual institutions. Last year, Kibbe and Nadaf launched Cancer Informatics for Cancer Centers, a national nonprofit that holds in-person meetings and conference calls on topics such as intellectual capital and cloud computing.
Much like the group of nonprofit cancer organisation CIOs, each IT leader brings his or her own areas of expertise to the group. London, for example, shares his approach to delivering clinically actionable results. Kibbe would like to see the alliance go worldwide.
"If we want to dive deeper into solutions to our problems, we can't do it alone. That doesn't benefit the patient at all," Nadaf says. "The more we collaborate, the more of a difference we make."
The mission continues
Not every day in the cancer battle is a good day. Loss is a fact of life for CIOs in the field.
"One of the things I've learned from St. Jude patients is that while survival is incredibly important, what's equally important is to thrive with every day that we're given," says Machen on the morning he was to attend the memorial service for an 8-year-old patient who had become a dear friend.
"One of the most beautiful things in the world was to see how much she cared and loved others at a time in her life where she could have been self-centered. It's tough, but a day like today is an everlasting reminder of the awesome nature of our mission."
That patient focus is something Nadaf of UCSF picked up when he was starting out. He worked in the lab of internationally renowned lung cancer expert Dr. David Carbone.
"We had the very first gene therapy protocol in the country for advanced lung cancer," he recalls. "I was a young twentysomething and very excited." Nadaf would accompany Carbone on grand rounds and tumor boards weekly, and he saw patients suddenly slip away.
"I saw how fast patients died, and it was troubling to me," says Nadaf. "Today I know the difference can be in the data. There's power in the data. And my job is to continue to bring out as much as I can from the data to help each patient."
Given the prevalence of cancer, CIOs don't have to go far to see the impact of their work. "Everyone on my staff knows someone who's been affected. Some have fought cancer themselves," says Skarulis of Sloan Kettering.
"We're not doing something for some remote benefit, something that might do some good in the future. We see on a day-to-day basis how what we do effects people's lives." That creates a sense of urgency in IT. "Every day that we don't know something is a day we haven't helped someone," she says.
But that pressure to do the best possible work is self-imposed and ultimately a positive force, says Skarulis, pointing to the fact that Sloan Kettering's employee engagement numbers are "off the charts."
"Everyone who's engaged in this mission is enthused and infused with the knowledge that they're moving cancer care and treatment forward," says Kibbe.
Mission - it's a word you hear a lot when you're talking to CIOs in the cancer field. And, yes, you'd probably hear a commercial CIO throw the word around, too, and mean it. But the mission of IT leaders at these non-profits might sound odd to their for-profit counterparts.
"Our goal is to end cancer," says Ferro. "If we're out of business because we've solved the cancer problem, we'll find other jobs. We'd love to put ourselves out of business."
Cancer research may be more affordable in the cloud
There's too much data for small labs to handle affordably. The answer may be putting petabytes in a central repository so more researchers have access.
As the volume of data generated by genomics-related cancer research technologies has grown, the storage, transmission and analysis of data has become too costly for individual labs and most small-to-midsize research organisations to handle.
So the National Cancer Institute (NCI) is looking to the cloud to make access to large, valuable data collections and advanced computational capacity available to as many doctors and scientists as possible.
NCI says that, during a two-year pilot program, it will award up to US$20 million to three cloud providers that can meet their technical and cost criteria. The ultimate goal is to build one or more clouds filled with data from the NCI's Cancer Genome Atlas and allow researchers to tap into it using data mining and analysis tools.
The Cancer Genome Atlas's current petabyte of data will grow, by the end of this year, to 2.5 petabytes of genomic information from 11,000 patients.
Building the infrastructure to store all that would cost a research institution at least US$2 million, according to Warren Kibbe, NCI's CIO and director of its Center for Biomedical Informatics and Information Technology, making the task cost-prohibitive for many small colleges and other research institutions.
By putting the data in the cloud, possibly using an on-demand pricing model, NCI may be able to expand the number of researchers working with the data, thereby speeding up genomics-based cancer research.
IBM's Watson trained to suggest the best cancer treatment
Researchers, physicians and analysts at Memorial Sloan Kettering Cancer Center have been training Watson, IBM's supercomputer, for more than a year to turn it into a decision-support tool that helps medical professionals choose the best treatment plans for individual cancer patients.
The goal is to improve quality of care for cancer patients no matter where they are located by giving their doctors access to the same up-to-date research as doctors at the nation's leading cancer centers.
"Here at Sloan Kettering, our people are all sub-specialists. If you treat lung cancer, that's all you treat and you become very good at it," says Patricia Skarulis, the institute's CIO.
"If you are doctor in a smaller community, you might see lung cancer one day, then breast cancer the next week, then a thoracic patient after that. It's hard to stay up-to-date on everything you need to know. We want to export our knowledge to the rest of the world through technology."
Ultimately, Watson will import data about a patient--pathology and radiology reports, vital statistics, initial consultation, and any other pertinent information about the patient's health--correlate that with the current research and protocols, and come back with suggestions.
Watson might tell the doctor that there's a 60 percent chance that treatment A is best for the patient and a 50 percent chance that treatment B is best. The supercomputer might also suggest additional tests that will enable it to offer better probabilities.
"Changes in technology such as cognitive computing are increasing our ability to go in and look at things we haven't been able to look at in the past," Skarulis says. "We're on the cusp of some of the most exciting changes in medicine."
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.