Join MySLAS Social

2016 SLAS Leadership Forum: Genome-Based Therapeutics – Are We Finally There?

"I really like the European Union's definition of precision medicine, which is the right therapy for the right patient at the right dose at the right time," said Jennifer Harris, director of drug discovery biology at the Genomics Institute of the Novartis Research Foundation (GNF). "It's not only about getting the right drug to the patient, it's also making sure that the drug is delivered in a way that's going to be safe and effective as well."

Harris was one of three panelists in the 2016 SLAS Leadership Forum, "Genome-Based Therapeutics – Are We Finally There?," at SLAS2016 in San Diego. The forum, moderated by SLAS Past President Daniel Sipes, director of advanced automation technologies at GNF, was an opportunity for senior-level executives and key opinion leaders in life sciences discovery and technology to discuss and debate this timely topic and learn from one another.

In addition to Harris and moderator Sipes, panelists included Matthew McGinniss, a board certified clinical molecular geneticist and executive director at Genoptix, a medical testing laboratory that works with cancer specimens looking for the presence of mutations and making therapy associations as needed; and Karen Gütekunst, vice president of diagnostic development at Illumina, a company that develops sequencers and chemistry used on next gen sequencing instruments. Harris leads the GNF drug discovery area focused on early disease biology discovery to target identification and optimization to the drug candidate.

The Current Landscape

The panelists and audience agreed that great strides have been taken to understand the human genome since first breaking onto the scene in 2001 with the publication of 90 percent of the sequence of the genome's three billion base pairs by the Human Genome Project. But the general consensus was that there is a long way to go before realizing all the benefits of personalized medicine.

"Certainly there's a high degree of interest in using the understanding of the genome to develop targeted therapies or combinations of therapies," Gütekunst said. "If any of you are in the oncology space and you went to ASCO (American Society of Clinical Oncology) last year, everybody was talking about the new immunotherapies for oncology. There is broad acceptance that the genome is important – we need to understand it and we need to use it to drive drug development, certainly in this space.

"The challenge is that while we have a lot of tools now to better understand the genome, there's so much information that is still to be developed," she continued. "It's about managing that balance. Do we have enough information now? In some cases we do. There are really great success stories – like the targeted therapies for melanoma and BRAF, lung and EGFR, RAS and colon cancer – but there's still a lot more to be developed. We're at that tipping point where the major cancer centers are all using this deep understanding of the genome to understand their patients, but outside of those cancer centers, there's still a lot of work to do. And there is a lot of work to do to improve the technology as well."

McGinniss added his perspective as an end user. "We do a lot of sequencing; there's a lot of demand," he said. "Solid tumors have been well established for many years as are the sets of mutations people are looking for to help drive therapy associations. The doctors want that information now – within a week or so – and they schedule patient visits based on getting those results. It's a real challenge to deliver actionable results within a week or 10-day period.

"In peripheral blood and bone marrow, you don't have a localized tumor, so you don't know the percent tumor nuclei that you do in a solid tumor," he continued. "But treating physicians are faced with making therapy decisions based on these results. Whether they're looking at AML or a myeloid gene panel, they're looking through these genes and making treatment decisions as we speak. It's pretty humbling to be part of this – trying to deliver actual results in a timely manner and dealing with the issues Karen alluded to. You've got whole genome or restricted gene sets you're looking at. How do you manage all these variants? How do you manage the data sets? How do you make that scalable and deliverable? It is a real challenge."

Pharmacogenomics, the study of the role of inherited and acquired genetic variation on drug response, has provided an explosion of new data according to Harris.

"It's important, right?" she asked the audience. "Sixty years ago scientists discovered that the selective toxicity of primaquine was due to a deficiency of glucose 6 phosphate dehydrogenase. This illustrated a very real impact of the genome on drug efficacy or toxicity. There are a number of good examples now, but as the other panelists and I have noted, there still is a large amount of information that just is not known.

"I think you have cancer in one category; we have molecular-targeted therapies absolutely, and we can go through some of those success stories," Harris continued. "I was lucky enough to be involved in one of those examples. But if you really think about the big, common diseases that we're still trying to tackle, a lot of patients don't respond to the therapies we have out there. We often don't know why patients don't respond in the way we think they should to the medications. I think it's more than just the genome. There are all the other layers that go on top of it – the environment, your diet, exercise, other medications. I think that is where we need to be moving to really have the promise of precision medicine and being able to treat a patient and have it be 100% effective."

Managing Expectations

"People need to understand the limitations of their tests, from the lowly morphology, flow cytometry, IHC (immunohistochemistry); you really have to be aware of the limitations of tests," McGinniss offered. "We have to be careful of some of those boundaries and understand, hey, this is the result of a hybrid capture assay using DNA; it's not using RNA. Or this is the result of an RNA-based hot spot mutation panel. We have to be very aware of that because a lot of the people that are consumers of this information, for example clinicians who get our test results, are not very aware of some of these limitations. It's our job, whether in research, clinical trials or routine diagnostics and therapy associations, to help people understand that you've used the best available test, but it does have limitations."

Consumers hear much about the potential of personalized medicine daily and for patients affected by diseases without cure, they may need guidance as to what this really means for them. In the United States, personalized medicine was elevated in patients' minds with President Barack Obama's 2015 State of the Union address announcement of the Precision Medicine Initiative. Obama called on Vice President Joe Biden to lead a new national $1 billion 'Moonshot' initiative to eliminate cancer as we know it by utilizing genomic analysis and enhanced data sharing among other areas.

These are huge promises. Most in the room agreed they expected effective genome-based therapies in their lifetimes.

"I'll put myself out there," said Gütekunst. "I do think that we're going to be able to do CRISPR-based therapies in our lifetime. The first entry will probably be cell-based type therapy where you're adding cells back in. The challenge does go back, though, to identifying what are you correcting? Or what are you putting in? What is that drug going to be? I think we all still have our work cut out for us in the drug discovery area in terms of finding new targets, finding new pathways to treat disease. Yes, we have more outlets – protein therapeutics, biologics, RNA, low molecular weight – but the fundamental problem of figuring out the biology and how that's actually going to impact the patient is still there."

"You have to manage expectations appropriately and you don't want to get people over excited about things," McGinniss cautioned. "I used to be in inherited genetics and it would just kill me when I'd go into a muscle disease clinic two days after there was a popular press article about some cure for Duchenne muscular dystrophy or something, and everybody's all hopped up and excited and you don't have much to give them. It's very frustrating."

Are We Doing Enough?

Several members of the audience queried whether the panelists thought those in drug discovery were using enough personalized information. Another twist on this comment was that given the acceleration in the technology, was data analysis keeping pace with technology development?

"From my perspective, I think that the technology is being widely adopted and widely used, so there's very strong interest in gathering information to better understand how to do drug development," Gütekunst said. "Has it actually impacted drug development? I think it's transitioning but I wouldn't say that that revolution we're all waiting for has happened yet. I think the biggest barrier to that is the informatics challenge and the lack of sharing information across major cancer centers, universities, government institutions or private companies like my own. The genome is so complex that we need many sources of information to contribute to that knowledge pool. We have to find a way to do that.

"Before I came to Illumina, I worked in a molecular pathology lab," she continued. "We used to say if you've seen one cancer, you've seen one cancer. They are so diverse. We need to get that critical mass of information so that we really can start to segment things further than population based. Everybody gets excited at Illumina when we talk about moving everything to the cloud. Wouldn't that be cool? But there are many challenges to that, such as placing personal health information in the cloud."

Several audience members shared how their organizations have committed to sharing information and forming extensive partnerships to do their part to accelerate discovery.

"There's no doubt that, certainly from the genome perspective, the Myriad case in the U.S. was certainly a huge step forward in the ability to not patent genes, not own naturally occurring information, even if you were the first person to identify it and discover it," Gütekunst said. "I think that's been huge in opening up the ability for people to pursue these genomic or genetic-based tests and information. It forces the intellectual property landscape more to the technology – owning the sequencing by synthesis chemistry, for example, or unique attributes of our sequencers that allow us to go faster, or less expensively.

"I think this is accelerating things a little bit," she continued. "But at the same time on the genomic side of things – bioinformatics and understanding how you do variant calling and how you process that information – I think there's still a lot of secret sauce around some of that information. We need the ability to open that up and share and help compare these things."

Gütekunst believes that the information exists, but is in many cases being siloed or protected. The infrastructure is not right.

"If you're in academics, you are rewarded for publishing novel information," she said. "And that doesn't necessarily encourage you to contribute to a bigger pool of information. If you are in industry in companies like mine, we make a lot of money on intellectual property, so our environment is structured around not sharing as much. Because of the complexity of the information that we need to understand, the only way to accelerate that understanding is through information sharing and incentivizing the community to do that."

McGinniss agreed. "If we can look for ways to incentivize some sharing with ways to preserve some of the aspects of privacy or intellectual property it will be very helpful," he said. "In the world I live in, we have a team of three people that look at variants every day. We report hundreds of next gen sequencing results for real patients every week. It's a challenge to manage that data, look at data, make assessments. We're in the trenches dealing with it every day."

Additional Conversation Threads

With robust interaction between panelists and participants, several other areas of interest were discussed during the 2016 SLAS Leadership Forum.

Informatics Talent
There was concern expressed time and again during the open discussion that with all of the data currently being shared and that which the group hopes would be shared going forward, there is a shortage of talent available with informatics expertise.

"I think the free market will take care of this issue," McGinniss said. "If you look at the salaries of people who have those skill sets, they're pretty well valued. We have trouble recruiting and retaining them, and I think that to some degree, speaks for itself. They're valuable people, and I think there are increased numbers of programs and academic programs training them."

Genomics for Disease Prevention
Participant Marcie Glicksman, vice president of biology, Orig3n, asked, "since prevention is actually the most effective treatment for a disease, what were the panelists' thoughts about using genomics and other tools to elevate prevention?"

"The strongest predictor of your long-term health is looking at your parents, your family history," Gütekunst said. "And yet, in the U.S., obesity is a huge issue. People still smoke. If you did a genetic test, however, and said, 'Oh, you have a predisposition for cardiovascular disease,' suddenly that seems more factual to people. So it is possible that the implementation of genetic testing could psychologically motivate people to change their behavior. Knowing your parents died of heart disease, you probably shouldn't smoke and you should probably lose weight and exercise. Give them a lab report that says you have a higher probability of developing cardiovascular disease does seem to be a higher motivator to change behavior."

Ensuring Quality
"We've talked a lot about data and the importance of the integration of that data," offered another audience member, Sue Crimmin, vice president discovery supply at GSK (GlaxoSmithKline). "But the data you generate to inform your hypothesis is really dependent on the quality of the material that you're analyzing. What are we doing as an organization, a community, to ensure that the material that we're collecting and the specimens we use have high fidelity and are relevant to the disease state?"

"We need to be aware not just of the test platform for reagents but of the limitations of cell lines," McGinniss said. "They're great tools and as Michael (Gottesman, SLAS2016 keynote presenter) pointed out in his talk today, there are great opportunities there. But you have to be aware of the bias of ascertainment, of developing the cell lines and the biases of making things grow in a nice, little, warm culture with certain growth factors and certain levels of CO2 and oxygen. We need to be aware of our choice of doing research grade sequencing or clinical grade sequencing because you're going to see differences."

Editor's Note: Michael Gottesman's SLAS2016 keynote presentation, "Analyzing the Complexity of Drug Resistance in Cancer," is available on demand for free. 

The Knowledgeable Consumer
Another discussion thread involved information available to the consumer and how that can push academia and industry forward.

"I think the patient advocacy groups have been really good at bringing data together and making people share," Harris offered. "At least what I've seen is that they encourage you to share the data, to really understand what the commonalities are and not to be in that competitive space."

Similarly, participant Jeff Paslay, principal, Paslay Consulting mentioned his experience when attending a local science program on one's personal genome. He indicated the crowd – some 200 or so, mostly retirees – were asked to raise their hand if they had sent a sample off to 23andMe. About half had done so and either had already or were planning to take their information to their doctors. The participant wondered how our industry was educating medical providers to deal with this new level of information.

McGinniss believes professional medical organizations were helping educate their members but it was the usual "technology always moves quicker and it just takes a little while to catch up as a culture and adjust accordingly."

Harris mentioned an innovative program in a Boston hospital where a medical geneticist helps physicians learn to address these questions. They role-play and videotape the doctor responding to patient questions about risk factors learned through genetic testing. "Then, they go back and figure out where the mistakes were made and try it again."

Closing Thoughts

Moderator Sipes is grateful to panelists Gütekunst, Harris and McGinniss for bringing their expertise and insight, as well as to the audience for their lively contributions to the discussion. After the SLAS Leadership Forum and throughout SLAS2016, participants continued to approach Sipes and resume discussion topics. "This is surely a sign, as the panelists suggested and Dr. Gottesman emphasized, that the use of patient sequencing information to guide genome-based therapies is still developing and the impact will be profound," Sipes concludes.

Images courtesy of the U.S. National Institutes of Health National Human Genome Research Institute.

April 11, 2016