IBM Watson for Genomics: The eHarmony for Cancer Treatment
At Intouch, we get pretty excited when someone or something comes along with a whole new way to hack healthcare for patients, and this week, the kudos go to Quest Diagnostics and IBM Watson Health.
On Tuesday, Quest and Watson, in collaboration with leading cancer institutions around the United States, announced a partnership called “IBM Watson Genomics” that pairs genomic sequencing and the power of cognitive computing to provide more precise treatments for cancer.
Watson may be most famous for winning the million-dollar jackpot on the game show Jeopardy in 2011. But the supercomputer was put to more meaningful use in 2013, when it was first used to help develop a treatment plan for lung cancer patients at Memorial Sloan Kettering Center (one of the medical centers collaborating on this project). At that time, Watson developed treatment plans based on existing medical evidence, patient case files and medical training cases. Combining Watson’s computing power with genomic sequencing takes its potential to help patients to a whole new level.
How It Works
The first step toward initiating an individualized cancer treatment plan involves a biopsy of the patient’s tumor. The sample is sent to Quest, where it’s sequenced and analyzed to identify mutations in the tumor’s genomic makeup. The sequencing results are fed into Watson, which IBM says, “will then use the sequenced genetic data and compare those data against massive bodies of clinical, scientific and pharmacological databases to help uncover potential therapeutic options that match the patient's tumor mutations.” A pathologist then reviews the results and sends them to the physician.
Essentially, the service can be compared to an eHarmony algorithm — one that matches cancer patients with the most promising treatments for their specific cancer.
Hope for Patients
The Quest-Watson partnership opens doors for many patients that until now have been all but shut. Because Watson can do the heavy lifting by rapidly searching its databases, even patients living in rural communities with limited access to specialists will have a shot at treatments that may have previously only been accessible to those in urban areas.
What Does It Mean for Pharma?
At first glance, it would seem there’s potential for Watson to more quickly match patients to clinical trials and established treatments so that they can begin to get well sooner. For now, there are more questions than answers:
- What data is Watson using, specifically, and who chooses what’s fed into its database?
- How is "best-course" defined?
- How will frequently used first-line treatments be weighted against newer entrants?
- How can pharma help Watson see the full benefits of a drug, especially if data is limited?
- Will Watson’s aggregated results and recommendations be made available to the public?
- Which oncology treatments may benefit from Watson’s recommendations, and which ones may see a drop in use?
- Will payers argue with Watson?
- Will all oncologists accept the idea of a computer making treatment recommendations for them?
- Is "Most frequently recommended by Watson" a convincing claim of the future?
We’ll be keeping tabs on this game-changing approach to cancer treatment, and we look forward to celebrating its successes.