On Monday night, Sen. John McCain (R-Ariz.) announced that he would leave his recuperation from brain surgery and return to Washington for a critical vote to repeal the Affordable Care Act. This dramatic announcement underscores the mettle that McCain’s friends and foes noted when the diagnosis of his brain tumor was revealed last week. Surely a man who could survive unfathomable hardships in a POW camp had what it takes to beat cancer.
Unfortunately, McCain is facing long odds. Glioblastoma (GBM), the kind of tumor he has, is an aggressive and deadly cancer. Only 50 percent survive more than a year from diagnosis, and only 5 percent survive five years or more. Just in the realm of politics and just in recent memory, this type of tumor claimed the lives of Sen. Edward Kennedy and Delaware Attorney General (and the vice president’s son) Beau Biden.
Thanks to efforts such as the Cancer Genome Atlas, funded by the National Cancer Institute, the scientific community knows more about GBM than just about any other cancer. We have created massive databases of molecular information about GBM patients, allowing us to map the cancer’s genomic profile. On top of that, because of the rapid adoption of precision oncology, hospitals and cancer centers are gathering a wealth of real-world data about GBM patients, including how molecular data is tied to treatments and outcomes.
Yet this information remains siloed, with clinical information cordoned off in one system and the molecular information in another.
For us to get closer to a treatment for McCain and the thousands of patients that will come after him, physicians and researchers need to be able to easily share data. We need to tear down the wall between the data so that we can radically reduce the lag between insights and treatment. And for that to happen, those in the lab and at patients’ bedsides need McCain’s colleagues in government to step up and act right now to remove obstacles to data sharing.
There are two main challenges when it comes to working on a cure for GBM. First, GBM is rare: there are only about 12,000 cases a year in the United States. Compared to a more common cancer such as lung cancer, with 220,000 cases in the United States alone, there are just too few GBM cases to study.
Second, GBM is complex and adaptable: It is not clear what drives GBM to occur, grow and aggressively resist treatments that have worked in so many other cancers. To make matters worse, GBM is lethal even when it does not metastasize outside of the central nervous system.
With too few cases and an ever-changing foe, advances in treating GBM will come not from a traditional clinical trial but from the aggregation of data — including genomic, clinical, treatment, imaging and outcomes data — from real-life care across the country and beyond.
Recognizing this, a group of oncologists, neuroscientists, other researchers and patient advocates from around the word started GBM AGILE, an effort to share the data and knowledge about GBM and how to treat it. We know that gathering this real-world data and giving researchers and doctors access to it is the key to progress in treating GBM. Yet here in the United States, there are needless regulatory barriers to making this happen.
Of course, protecting patient data and privacy is an important concern in health care, and for the past two decades that has been governed by the Health Insurance Portability and Accountability Act of 1996.
Yet HIPAA never anticipated the modern advances in research and data; when it was written, most people still dialed up to access the Internet, if they had access at all.
The law is not clear as to what kind of data-sharing is permitted in routine clinical study. With this ambiguity, we have found time and again that hospital system attorneys counsel their doctors not to share aggregated data or risk a huge fine. The result is a culture of fear in which hospital systems err on the side of not sharing data rather than furthering scientific — and clinical — progress.
The Department of Health and Human Services can easily clarify these rules and make it clear that aggregated de-identified patient data — including patients’ molecular, imaging, treatment and outcomes information — can be shared between health care providers. Moreover, HHS should emphasize that molecular data about a patient’s tumor is part of the patient record, just like any other test result. Thus it cannot be subject to “data blocking” by health-care organizations such as testing labs, who may want to hoard the information for themselves.
Realistically, these changes most likely would not produce a result that can help McCain. But if the president and his administration want to honor his fight with GBM, they should welcome him back to Washington by immediately removing these barriers to data-sharing and allowing physicians across the world to have the data they need to make progress against this killer.