This year’s Precision Medicine World Conference (PMWC 2017 Silicon Valley) was comprised of many exciting sessions and presentations (almost 300 speakers across four tracks) in the area of next-generation sequencing, different diagnostics applications, precision medicine, big data analysis, the microbiome, large cohort studies, biobanking, and data interpretation/knowledge extraction. In addition to providing a great set of overview talks on latest developments and achievements across the health care sector, in pharma, and related to regulatory aspects, this latest rendition of PMWC also featured a government presence (the former FDA Commissioner Dr. Robert Califf and the former Cancer Moonshot Task Force Director Greg Simon) and Elizabeth Baca (Health Advisor to Governor Brown’s Office of Planning and Research) among others that shared with the audience their respective updates on the stand of the FDA on clinical testing, the “Cancer Moonshot Initiative”, and the “California Initiative to Advance Precision Medicine”. Furthermore, with personalized medicine establishing itself, new emerging themes were on the agenda, such as wellness, the microbiome, the CRISPR technology, new advancements in liquid biopsy and immunotherapy, or knowledge extraction and data sharing – the latter being a critical aspect in making personalized medicine a continuous success.
In this post, I am summarizing just a couple of personal conference highlights, admitting that I was not able to see all talks or visit all tracks, and thus cannot to do justice to a number of other high quality presentations that would have been equally worth covering. To read up on additional PMWC 2017 Silicon Valley content check out the BioQuick News blog Precision Medicine World Conference (PMWC) 2017 Winds Up in Silicon Valley, Themes of Immuno-Oncology & Scientific Wellness Dominate or Sanjay Joshi post The Patient’s Perspective: a critical technology driver.
Let’s start out with a set of interesting quotes by presenters:
- Allison Ballmer (Roche): “the reason genetics is not seeing greater success, because there is a lack of education”
- Euan Ashely (Stanford Healthcare): “We are no longer just at the dawn of clinical genome sequencing.”
- Georgy Tsongalis (Dartmouth College): “Very difficult to find people with NGS experience – ‘the biggest challenge of my career’. Some techs – can’t get trained.”
- Greg Simon (Cancer Moonshot): “”Data sharing isn’t the answer to everything but data hoarding is the answer to nothing.”
- Greg Simon (Cancer Moonshot): “determinants like location (where you live) and insurance are the biggest factors of cancer survivorship.”
- Jennifer Doudna (Berkeley): “Science Matters”…..highlighted in Women’s March”
- James Allison (MD Anderson Cancer Center): “It would be great to have predictive, prognostics, or pharmacokinetic markers for cancer immunotherapies”
- Jordan Shlain (Private Medical): “trying to sell something to someone who is anxious (a patient) is different than when you are trying to sell something to a person who is excited (a consumer)!”
- Kathy Giusti (Multiple Myeloma Research Foundation, MMRF): “Patient data is more important in the context of thousands of other people, but patients don’t know that.”
- Kathy Giusti (MMRF): “Data sharing incentives: access to genomes, data, money, and players.”
- Lincoln Nadauld (Intermountain Healthcare): “Three major hurdles of implementing precision medicine: 1) interpreting data, 2) drug procurement, and 3) payment.”
- Randy Scott (Invitae): “sequencing your genome is not an event, it’s a journey!”
- Robert Califf (FDA): “We need more transparent sharing of data, and we are not there”
- Robert Califf (former FDA Commissioner): “It’s critical to have a bedrock of safety – think about the supply chain of breakfast items”.
- Robert Nussbaum (Invitae): “In no other area of medicine (but genomics) do people actively “hide” data– we need clear ethical standards.”
Fireside chat with (now former) FDA Commissioner Dr. Robert Califf and Dr. Keith Yamamoto
This year’s conference was chaired by Prof. Dr. Keith Yamamoto, Professor UCSF, Vice Chancellor for Science Policy and Strategy, Vice Dean for Research, School of Medicine. Yamamoto kicked the main program off with a fireside chat with the (now former) FDA Commissioner Rober Califf, discussing the FDA’s Role and Responsibility in Expediting Best Treatments for Patients and how much they should be involved in discussions when it comes to safety. Califf mentioned that in the glorious days of the FDA the advisory committee was up to speed, but now the system has changed considerably. Medical reviews often are not best advised, and that is mostly due to the fact that not enough experts can be found. As it stands, the advisory board currently functioning is unfinished business. Interestingly, people with certain diseases are getting involved and are working with the review committees. Going forward we have to think about the process differently: it is not just anymore someone isolated sitting at the FDA – it has to be an interactive process. The impact of advocacy groups, doing it for the right reasons is immense. Califf pointed out the exemplary UCSF and Stanford collaboration, and that something similar would be ideal as an interaction between the FDA and the community. One other aspect/challenge he pointed out: academia “produces a lot” with the purpose of clinical discovery, but not always with reproducibility in mind. This has to change! Lastly, Califf did emphasize that data hoarding is still a big problem and we need more transparent sharing of data.
Califf also mentioned that when it comes to new technology, we are at an exciting intersection of development, innovation, and application. Innovation has to occur, and product development has to keep pace with it. We are not there yet. There are many software errors that need to be straightened out, same for LDTs. We have to come up with a common ground. This should be an up- and coming focus for 2017.
The Value of Data Sharing
This timely panel had all the right participants: Lincoln Nadauld (Intermountain Healthcare), John Leite (llumina), Sharon Terry (Genetic Alliance), John Shon (Illumina), Juli Klemm (NCI), Kathy Giusti (MMRF), and Robert Nussbaum (Invitae).
The panel started off its discussion with an interesting survey previously performed by Illumina with a focus on the data sharing challenges . Interestingly, very high on the list besides lack of resources, motivation, and data standardization, is a lack of business incentive to share data (see image below).
Kathy Giusti pointed out that when dealing with large data and trying to solve a problem, it is important to ask the right questions and this includes payor questions. One of the struggles is knowing the landscape of the data. What that means is we need fewer and bigger data islands, but we can’t do that without standards. We need a federated model and standardized data sets. Additional points made by Giusti:
- The patient will eventually drive the data sharing disruption.
- Patient data is more important in the context of thousands of other people, but patients don’t know that.
- We need to bring teams together and make them work together and share their data.
- It turns out that organizations that have data do not work with payors because they do not know them.
- Another problem is education: each data set requires education when new teams come in to join the analysis.
- What we need to do is to optimize patient care – there is a space for everybody! Eventually everybody in the space will benefit including the payor.
- Patients get creative when they get sick. One thing we learned: don’t ask for the genome but ask about a specific marker via social media (e.g. Facebook) and it turns out patients will start sharing some of that information.
Sharon Terry emphasized that it keeps coming down to the business concepts: making data a commodity. But we have not yet aligned ourselves around key question “what are the people’s needs?” Clearly, the incentives are not aligned. She considers it a huge uphill battle. If the data was openly accessible, we could transform the system. Additional points made by Terry:
- The revolution is up to the people – the point is that the patient is the answer and we do not let them know that – we need to make it as easier and more transparent like a transaction (think paypal).
- If we give people that power they will be on a march together. We have two classes of people and if done correctly they will empower each other: the advocate and the sick one.
- Looking beyond clinical data, only about 50% of academic research centers share data and results, and if they do, the data is often not complete.
- Sharon suggested to even going as far as removing an IRB if the researcher is not sharing the data.
Lincoln Nadauld put some of the emphasis on the huge value in data sharing, but also pointed out that there are still huge roadblocks, such as interoperability, regulation, and money. Additional points made by Nadauld:
- Private companies have profited from keeping private data. We want and need to break this down.
- Data is valuable, but it is only valuable if you keep breaking it down into the context of the data – the raw core data should be shared.
- The data generator should be responsible to share the data and should not just generate it.
- Stakeholders have to get together and solve it together so it is not regulated by a third party.
- Patients are the common thread.
- We need standards and we have to demonstrate that data sharing is valuable.
- We should engage with the patients via social media so others will sway into that same direction.
Robert Nussbaum commented that sharing data for the purpose of clinical care and research is important. Additional points made by Nussbaum:
- The data ultimately belongs to the patient and it should be used to help them/advance their diagnosis or treatment. Patients want the full view.
- There is structure in genetic/genomic data and that information is available.
- At a simple level there is no question that data sharing is valuable
Julie Klemm pointed out that broad data sharing is important and it does not end with the data generation. Additional points made by Klemm:
- Data is not the commodity.
- When the government mandates data sharing without incentives there is no motivation to do so.
- We need to empower the patient.
Walking on the moon
It was great to get a first-hand update on the Cancer Moonshot Initiative and its goals from Greg Simon, the (now former) White House Task Force Executive Director for the Cancer Moonshot Initiative. Clearly, the biggest delivery achieved is the fact that 20 different agencies (e.g. DOE, NCI, DOD, etc.) started to work actively together to tackle the issue and to overcome the assumption that to achieve something it will take a long time, especially in science. This was achieved by hiring the best and smartest people, by talking to as many people on a random basis as possible, and by engaging with tools such as social media to turn things around (e.g. started an anti-tobacco campaign). A lot was learned about the issue of cancer treatment and what the determinants for survivorship are: where you live and what insurance you have determines more than any other variable (even family history) or race and economic status whether you survive cancer.
Simon also pointed out that the field needs to tighten the spaced between data and organizations that can improve health.
During the nine month cancer moonshot period, quite a bit was achieved as summarized in an earlier PMWC blog. End of June 2016 the task force held a workshop with 400 people, and since then over 270 different workshops in different locations have taken place – clearly an indicating that the cancer moonshot that started as a government program is now a movement. Nice to see that a cultural change has occurred that was instilled by Vice President Biden’s sense of urgency.
The $100 Genome Paradigm
Clearly 2017 is all about pushing the price per genome down even more – we already heard it from Illumina at this year’s JP Morgan Healthcare Conference in San Francisco, and yet again at this year’s PMWC 2017 Silicon Valley conference from BGI. Though BIG’s goal is to deliver a $100 (-$200) genome by 2020, this year they expect to bring the cost down to about $600. It is not clear whether this price is just for the sequencing part or whether it also includes data interpretation and reporting.
Yiwu He also introduced the China Precision Medicine Alliance (involving >50M outpatients across different cancer hospitals and university affiliated hospitals in Jiangsu, Shenzhen, and Guanzhau), the different tests BGI has conducted (e.g. cancer, HPV, or NIPT tests), and some details on the China National Gene Bank which includes collected human, plant, animal and microbe samples.
Genetic testing in healthy populations
Randy Scott (Invitae) chaired this exciting session with speakers that included Robert Green (Genomes2People), Jordan Shlain (Private Medical), and Just Kao (Helix). This session started off by Scott making an analogy to the history and adoption of cell phones: it took nearly two decades for cell phones to take off and he sees an analogy to what is happening right now in the field of genome sequencing: within 20 years the number of “genome subscribers” will rival the number of cell phone subscribers”.
Scott made the case that simple rare genetic conditions are actually common in aggregate, and a ton of money is spent in healthcare without knowing the hereditary condition. Clearly, the question is what type of genome sequencing should a healthy person perform, considering that genome sequencing at 30/40x has still many “holes”. Scott pointed out that we are encountering “the last mile problem”, his main current motivation behind starting a new company (together with Robert Green) that addresses exactly this aspect. The company is called Genome Medical, which really is a medical company, with a medical practice, staffed with physicians, and medical care. The sole purpose is to help understand the data and help with the interpretation of it.
On the other hand, Robert Green discussed the MedSeq project which integrates whole genome sequencing into clinical medicine. It’s a unique randomized trial of whole genome sequencing in clinical care with adults and newborn babies, following people through many different models. The project had to overcome the perceived notion that genomic information is inherently harmful, bad, and distrustful, in order to successfully enroll patients. They learned that the type of people who want to find out about themselves via genome sequencing, are actually well suited and informed about managing the data and findings. Green pointed out that when it comes to genome sequencing, they of course run into the same issues around incidental findings. Therefore, establishing a standard for incidental findings seems important. They labeled it opportunistic screening – dual analogy (doctor tells you about a spot on the lung identified, after a whole a body scan is performed). With the MedSeq project they identified many of carriers within the general population – many carry a dominant trait, but no phenotype. They charge an extra $350 for the additional investigation of limited incidental findings. Green also mentioned the BabySeq project which is just about getting started. They reached out to 2,190 families with ~190 already consented to join the project. The reasons cited for low participation included concerns about insurance discrimination, and confidentiality as well as privacy concerns.
To sum it up, Green did say that sequencing in healthy individuals is still controversial, as unanticipated findings will be found, but it is clear that the demand may lag with societal attention to privacy.
Jordan Shlain, a primary care physician, called himself the voice of reality. He pointed out that in his mind primary care is pretty much a broken and a fractured non-system. In medicine, we tend to move slowly with the main goal not to kill people.
He also pointed out that there is a difference between a consumer and a patient, and that possibly a third category (not yet defined) is needed, especially when it comes to genetic testing. His perspective: trying to sell something to someone who is anxious (a patient) is different compared to when you are trying to sell something to a person who is excited (a consumer)!
Shlain proposes the Trust-ome – in order to get something done in medicine: “change actionability to better prevent” AND “in order to get something new done you need trust”. Not only is it all about relationships, there also needs to be a thoughtful process in place to roll it all out!
Last but not least, we heard Justin Kao present the latest on the Helix platform which he proposes will bring genomics into everybody’s life. Kao suggested that in the next 5 years millions will have their genome sequenced, and that there will be a strong consumer interest. To achieve this, Helix has built CLIA certified labs, partnered with software companies, and launched with partners’ activities to build appropriate apps on an iOS platform available to everybody. The customer can purchase DNA sequencing service via Helix, with Helix handling the sequencing and providing access to DNA data via partner APIs and Apps. Helix’s solution will focus on data storage, privacy, and security. The long-term vision includes customized DNA-based nutrition recommendations for clients, encouraging specific behaviors, possibly extend to fitness, health, carrier screening, and much more.