7

Patient Perspectives

PERSONALIZED MEDICINE COALITION

Daryl Pritchard, senior vice president for science policy at the Personalized Medicine Coalition, invited workshop participants to change gears and focus on “the most important stakeholder in this process—the patient.” Pritchard started by drawing an analogy between precision medicine and modern aviation. He explained that traditional medical treatment—involving trial and error and symptom-based taxonomy—is akin to early aviation, when pilots had few tools to guide them and “basically looked out the side of the cockpit for signs that they were going in the right direction.” In con-

trast, precision medicine and modern aviation involve the use of complex and precise technologies that help guide the provider or the pilot to the right treatment or destination. Expanding this analogy further, Pritchard said that each of the instruments of precision medicine is analogous to a step in flying a plane. Susceptibility biomarkers—which can help predict the risk of developing a disease and indicate the need for increased surveillance or prevention strategies—are similar to flight planning. Mechanistic biomarkers—which can help select the right treatment by indicating that the candidate drug occupies the intended target—are akin to a pilot’s navigation system. Finally, pharmacogenetic biomarkers—which can help select the right dose and duration by indicating dose-dependent target modulation and/or pathway inhibition—correspond to a plane’s landing instrumentation. These biomarkers “make sure that you get to that destination without an adverse event.”

Pritchard turned to the advances in personalized medicine over the last decade or so. In 2005, the cost to sequence a single human genome was $100 million. Today, said Pritchard, the price is around $1,000. In 2005, there were 14 personalized medicines on the market, and today there are more than 160, with many more in development. Pritchard reported on a study commissioned by the Personalized Medicine Coalition with Tufts Medical Center that showed that 42% of all the drugs in the pharmaceutical pipeline in 2015 were personalized medicines, and a full 73% of oncology drugs in development were personalized (Tufts Center for the Study of Drug Development, 2015). In concert with the availability of personalized drugs, the market for genetic testing products has expanded dramatically. As of 2017, more than 65,000 genetic testing products were on the market (Concert Genetics, 2017). In short, said Pritchard, we are in an era of unprecedented discovery.

Noting the variety of new major research projects around the world, including the 100,000 Genomes project and the All of Us research program, Pritchard said that these large-scale human genome databases are an ambitious start that will need to be supplemented with research about mechanisms and molecular pathways of disease, and clinical relevance of variants will need to be validated through animal-based and in vitro models. Perhaps most importantly, findings need to be reproducible in order to be translatable. Pritchard imagined the process like a loop where findings from large database studies feed into animal-based and in vitro research, which in turn feed back into analysis of the large database, and eventually into translation to the clinic.

Of the key stakeholders in precision medicine, researchers, regulators, and the diagnostic and drug industries are on board. However, the stakeholders that are most critical for moving precision medicine into the clinic—payers, providers, and patients—are more cautious, remarked Pritchard.

Patients, in particular, need more information and more dialogue regarding the value of precision medicine. The Personalized Medicine Coalition performed a survey in 2014 and found that 62% of those surveyed had never even heard of personalized medicine, said Pritchard. Only 10% of people said that their doctor had discussed personalized medicine with them or recommended a targeted treatment. However, the survey also found that, when participants were told of the benefits of personalized medicine, they were quick to see how it could help them as patients.

Pritchard closed with two narratives. First, he talked about the progress made on leukemia and lymphoma in the last 60 years. In the 1950s, these diseases were simply referred to as diseases of the blood. In the 1970s, they had been differentiated into a few categories. Today, there are nearly 90 different subcategories of leukemia and lymphoma, each with its own optimized care pathway for effective treatment, said Pritchard. The effects of this have been remarkable, with the 5-year survival rate climbing from about 0% in the 1950s to about 70% today. Second, Pritchard told the story of a patient, Deb Smith, who was diagnosed with advanced non-small-cell lung cancer 5 years ago. She and her doctors had not expected her to survive long, so she had focused on spending her remaining time with her son so he could have memories of his mother. She was fortunate to be enrolled in two different clinical trials and is alive and well today, saying, “I’ve been taking trips with my son that I never thought I would be able to take . . . what I’ve been trying to do is give him some good memories . . . and in the process, I’ve found that my life is more complete than ever before.” These stories, said Pritchard, demonstrate that the impact of personalized medicine is patient centered and is profound.

PARKINSON’S DISEASE

The Michael J. Fox Foundation for Parkinson’s Research (MJFF), said Brian Fiske, senior vice president for research programs, is dedicated to finding a cure for Parkinson’s disease through an aggressively funded research agenda, and to ensuring the development of improved therapies for those living with Parkinson’s today. MJFF has invested nearly $750 million in research programs to date, funding both domestic and international projects in both the non-profit and for-profit sectors. MJFF funds and coordinates research in two areas that are particularly germane to animal modeling for precision medicine. First, MJFF works to develop tools for researchers, including new preclinical models. Second, MJFF supports the collection of extensive clinical and biological data from patients, who, Fiske notes, are the best models of Parkinson’s Disease. Both the tools and the data are made available to the research community.

Parkinson’s disease is an ideal target for a precision medicine approach,

said Fiske. People with Parkinson’s vary greatly in progression, symptom severity, and response to treatment, suggesting that there are potential subtypes of the disease. The cause is thought to be multifactorial, with both identified and unidentified genetic forms as well as many idiopathic cases. The biological understanding of the disease suggests that many cellular systems are involved, including mitochondrial function, cellular protein handling pathways, and lysosome biology. It is unclear at this stage whether or how these pathways intersect or interact with one another, but it is clear that there are multiple mechanisms involved, said Fiske. As Parkinson’s shares some similarities with other neurodegenerative diseases, whether these diseases should be reclassified based on their biology and mechanisms rather than clinical description remains an open question.

Recalling Hook-Barnard’s presentation that compared human data to the layers of a GIS map, Fiske said that, in order to fully understand and thus effectively model Parkinson’s, there should be a map that includes data from molecular information all the way to the patient’s daily experience with the disease. Understanding each of these layers, and how they interact with one another, will lead to a clearer understanding of the disease and its subtypes. Fiske noted that the description and classification of Parkinson’s has not changed significantly since it was first identified 200 years ago. The disease is still largely diagnosed on the basis of motor symptoms, such as bradykinesia, rigidity, resting tremors, and postural instability, despite the fact that there are a number of non-motor symptoms associated with Parkinson’s, such as cognitive impairment, mood disorders, autonomic dysfunction, and REM sleep behavior disorder.

While the diagnosis of Parkinson’s has not changed significantly, research on the underlying pathology has changed the understanding of the disease and the work on potential treatments. Some of the symptoms of Parkinson’s—primarily motor—are associated with the loss of dopaminergic neurons in the substantia nigra and the presence of Lewy bodies (abnormal aggregates of the protein alpha-synuclein) in the brain. These criteria are used for neuropathological confirmation of the disease, and have been the primary drivers for development of potential therapies for Parkinson’s. However, a number of other pathologies as well as brain cell degeneration have been observed in Parkinson’s patients, suggesting that the classification of Parkinson’s is more complex than a simple description of symptoms and pathology.

Parkinson’s is a progressive disease that changes over time, said Fiske. The early signs of the disease tend to be largely motoric, the progression of which can be somewhat slowed with currently available treatments. As the disease progresses, the non-motor symptoms become more prevalent. In order to make better treatments for people with Parkinson’s, it is important that models can reflect all stages and symptoms of Parkinson’s, said Fiske.

However, even understanding or classifying the clinical differences seen in patients is challenging. Patients are often classified into two clinical subtypes of Parkinson’s: tremor dominant or postural instability gait disorder (PIGD). A large prospective cohort study funded by MJFF, however, found that these classifications were not particularly stable over time (Simuni et al., 2016). Patients who had been classified as tremor dominant at baseline (i.e., when newly diagnosed) tended to largely stay that way after 1 year; however, a significant number of patients who had been classified as PIGD had switched to tremor dominant after a year. Fiske said that whether these clinical differences represent true disease subtypes or just snapshots of the same disease over time remains a major question in the field.

A number of animal models have been developed to study Parkinson’s, but to date none are considered predictive of the human condition, said Fiske. These models are divided into two categories: those that model environmental triggers, such as neurotoxins or inflammation, and those that model genetic triggers using strategies such as knockouts or transgenics. One of the most critical needs in Parkinson’s research, said Fiske, is finding a way to accurately measure alpha-synuclein in the brain. Because there is currently no way to image the protein in the brain, attempts have been made to measure it elsewhere (e.g., spinal fluid) and use those data as biomarkers for Parkinson’s. Animal models will be essential to this work. One model in particular is based on the idea that alpha-synuclein pathology might physically spread from cell to cell in the brains of people with Parkinson’s. The model meets some face and construct validity criteria for the human disease, but its predictive validity is still uncertain and debate remains in the field about whether this spread is a true feature in human Parkinson’s disease.

Fiske concluded that, while models can be extremely useful for understanding more about Parkinson’s, ultimately the best model of Parkinson’s disease is people with the disease. With this in mind, the MJJF has spent much time and effort on a major project called the Parkinson’s Progression Marker Initiative (PPMI), which collects data from cohorts of patients over time in order to build a clearer picture of disease and inform modeling. An emerging online version of the PPMI, called Fox Incite, uses a variety of emerging technologies and collection tools, such as wearables, to collect as much data as possible in order to better understand what Parkinson’s is so it can be better treated.