Call it the “too much knowledge” paradox: The vast amount of new data generated by biomedical science carries the promises to make medicine über-precise. In the meantime, however, it does so at the peril of generating too much noise and confusion.
As the still young field of precision medicine goes through its growing pains, scientists, clinicians and researchers are trying to determine how to harness rich, new data to ensure the information is interpreted, contextualized and, above all, used in a manner that renders medicine properly individualized for each patient? At the same time, those pioneering this field must learn how to turn all the emerging knowledge into workable clinical and business models.
These and other coming-of-age questions were addressed by patients, scientists and entrepreneurs at the third annual Precision Medicine Symposium, organized by the Department of Biomedical Informatics and held at Harvard Medical School on June 21. The daylong conversation explored three central themes: early disruptors in precision medicine, academic and industry collaborations, and guidance for consumer decision-making.
Some of the most critical insights into the field’s future, according to event organizer Isaac Kohane, chair of the Department of Biomedical Informatics at HMS, will come from harnessing the drive and passion of motivated patients and families to disrupt classic models of care delivery—from capturing the cross-pollination across academia and industry and from ensuring that patients and consumers are equipped to navigate the dangers of too much information and information “exhaust” from the data tailpipe.
Kohane thinks that the people most likely to move precision medicine forward are not necessarily the traditional bearers of knowledge—scientists, clinicians, researchers—but the patients and families who are profoundly, and all too often agonizingly, motivated to shake up the traditional ways of practicing medicine. These are the people Kohane likes to call “activated” patients, who use the urgency of their own fate to propel science forward.
Keynote speaker Shirley Pepke exemplifies the notion. In 2013, Pepke was diagnosed with late-stage ovarian cancer. The odds were bleak—44 months median survival.
Propelled by her instinct to survive and armed with a doctorate in physics and experience as a data analyst, Pepke dove into research and took on her own cancer.
She had her tumor’s DNA and RNA analyzed. Making sense of the DNA results was relatively straightforward, but the RNA profile turned up vast amounts of difficult-to-understand information. The epigenetic noise, Pepke says, was overwhelming. To make sense of it, she turned to a machine-learning expert at Caltech who specialized in detecting distinct patterns hidden within complex data sets.
Meanwhile, Pepke’s initial round of treatments kept the disease at bay, but in 2015 the cancer returned. Her oncologists started using words like “chronic,” “palliative” and “incurable.” Although she tried the standard, highly toxic chemo regimen recommended for a recurrent disease like hers, the side effects were profound, and her disease was not responding all that well to the treatment either.
Still, she persisted. Pepke came across a paper on a form of immunotherapy using the so-called checkpoint inhibitors—drugs that take the breaks off the immune system and unleash it to eradicate cancer. That research, however, had been conducted in mice. Pepke was well aware that what works in mice often fails in humans. Yet, she had little to lose.
Pepke took the paper to her oncologist who recommended combining the checkpoint inhibiting agent with stereotactic surgery to the abdomen. She was treated with a checkpoint inhibitor and underwent debulking surgery in the spring of 2015, which was followed by a round of chemo. Initially, her disease raged on. Pepke’s oncologist suspected that her kidneys—studded with and overtaxed by metastatic tumors—were beginning to fail. But a few months later, things took a different turn.
In the fall of 2015, Pepke’s tumor markers plummeted—by nearly half. They dipped even further during the winter, reaching their lowest values in February of 2016—a sign that her immune system was attacking the tumors.
Today, Pepke’s illness remains in remission. She is using the painful insights of her own experience to develop a tool that uses genome and exome data along with machine learning to build an algorithm that can profile each patient’s tumor signature to devise precision-targeted treatments. She is also on a quest to develop publicly available ovarian tumor data for analysis and profiling.
Pepke is living proof that highly motivated patient-citizens can catalyze change in the medical establishment.
“The people taking the reins are citizen-scientists, and they are changing our concept of who can be leaders in changing the course of human disease,” said George Q. Daley, dean of Harvard Medical School, at the symposium.
Though originally incentivized by their own personal crises, Daley added, these people are launching new business models because they are motivated by the need to spread knowledge and help others.
But creating successful business models that can disseminate knowledge and provide access to such transformative information is key to democratizing precision medicine and propelling it to a more mature phase, Kohane said.
To help this happen, he added, collaborations between academia and industry will be vital.
Academia and industry: playing to win
Some of the more disruptive ideas in precision medicine have come out of academia, while others have sprung up from purely commercial initiatives, Kohane said. When these two wellsprings of innovation find a way to collaborate, the results can be spectacular.
Take for example, Kathy Giusti, a pharma executive who, in 1996, was diagnosed with multiple myeloma. At the time, research on this disease was anemic, treatment options were few and life expectancy was a mere three years. Propelled by her diagnosis, more than 20 years ago, Giusti and her twin sister launched the Multiple Myeloma Research Foundation (MMFR), which for more than two decades has fused the acumen of academia to the nimbleness of industry. Since its inception, the foundation has helped catalyze the development of a series of drugs for the treatment of this rare blood cancer. These drugs have played a role in boosting the life expectancy of patients with the disease nearly threefold. Giusti hopes the MMRF model can be used as a template of collaboration between academia and industry to accelerate the pace of discovery and translation into clinical therapies for other forms of cancer and, indeed, other diseases.
“Our model brings all the players together to move the ecosystem forward,” she said.
It’s a sentiment echoed by others.
Citizen science and institutional science work better together than alone, said Jessica Richman, co-founder and CEO of Ubiome, a company that specializes in profiling individual microbiomes and develops big data analytics to inform science.
Yet despite all this new knowledge, the unknowns remain more than the knowns.
The patient: how to make sense of it all
By now it’s become a near-classic conundrum in medicine: A patient undergoes a genetic test for several types of inherited disorders. The tests results indicate variants of uncertain significance. Should the patient pursue more aggressive testing? Do these variants portend disease or are they benign, clinically meaningless variations of a normally functioning gene? More to the point, are all genomic tests created equal and are they equally calibrated to detect perturbations in one’s genetic make-up?
The sheer volume of data generated is overwhelming—a threat and opportunity scenario common in many emerging fields, the symposium participants agreed.
Understanding what information is clinically meaningful and actionable can be daunting for physicians and patients alike.
Kohane urged those working in the field of precision medicine to press forward and put big data in context that helps guide clinical and patient decision-making to avoid “potholes” along the way. Kohane envisions a “consumer reports” for genomic data that ensures physicians and patients are equipped with knowledge to make the right decision about treatment and care.
“We now have Consumer Reports that allows us to compare dishwashers, cars, tooth brushes. We don’t have anything like this in precision medicine,” Kohane said. “We have to be wary of not committing similar errors that other industries suffered from.”