No company has made a larger bet on using human genetics to drive R&D than Amgen. Here’s an update on the company’s quest to focus its research on targets validated in humans.
It was roughly five years ago that Amgen first announced its plans to aggressively leverage human genetics in its research. At that point, Amgen’s belief that genetics could transform the search for new medicines had already led to several major decisions and investments. They included the acquisition of deCODE Genetics, an Iceland-based company with an unrivalled gene discovery platform. Amgen had also revised its R&D strategy to focus on drug targets validated by genetic or other compelling human evidence. With the help of deCODE’s database linking variants in the human genome with disease risk, a number of programs that fell short of the more rigorous standards were jettisoned.
In the intervening years, Amgen has been conducting what amounts to one of the largest and most ambitious research experiments in the history of the industry. The goal is to show that human genetic validation of drug targets, applied as a forward-looking strategy, can help to cut the high failure rate that has plagued drug development programs for decades. The brutal attrition underscores the huge gaps that remain in our knowledge of human biology. Amgen believes human genetics offers a key approach to addressing both the complexity of biology and the industry’s R&D productivity problem.
“What we’re doing with deCODE is a profound contribution to answering basic questions in human biology,” said Sean Harper, Amgen’s executive vice president for R&D. “How do variations in DNA affect disease risk in humans? We’re learning more about that at deCODE right now than anywhere else in the world.”
Solving the Complexity of Biology
Amgen is also learning that success in applying genetics requires skills beyond the ability to identify disease genes. The biology that genetics reveals may be relevant to disease but it’s seldom easy, said Philip Tagari, Amgen’s VP for Therapeutic Discovery.
“Companies have always preferred to work on tractable targets, so most of the low-hanging fruit that happens to have genetic support has already been picked,” Tagari said. “That’s why the new targets emerging from genetics are often harder than the targets pursued by the industry in the past.”
Despite challenges, the rationale for Amgen’s large bet on genetics has only grown stronger since the strategy was adopted. The tools used to probe a gene’s function and alter that function have grown more sophisticated. Additional studies have bolstered the proposition Amgen is testing—namely that human genetic validation increases a medicine’s likelihood of success. In keeping with that conviction, the share of Amgen’s pipeline backed by genetics has climbed significantly in the past five years and includes a large number of programs based directly on insights from deCODE.
Improving Success in Clinical Trials
The first of these programs to enter the early stages of development targets a cell surface protein in the liver called ASGR1. A rare loss-of-function mutation in the gene that codes for this protein is associated with a 34 percent lower risk for heart disease. The mutation is intriguing because its cardiovascular impact is likely due to factors beyond the modest decrease it causes in LDL (bad) cholesterol.
“Inhibiting ASGR1 might provide additional benefits in patients who already have low levels of LDL,” said Simon Jackson, a scientific executive director who leads the drug discovery team. “We believe the potential is there to reduce cardiovascular events by a different mechanism.” Amgen has been advancing a number of therapeutic options to see which modality is best suited to the biology of this newly discovery pathway.
Leveraging two massive sets of data
Targets like ASGR1 show the usefulness of rare genetic variants. Unlike common variants typically found in genome-wide association studies, or GWAS, rare variants are more likely to be in regions of the genome that code for proteins.
The Value of Rare Variants
By dialing protein activity up or down, these rare mutations can have a more dramatic effect on disease risk and reveal important biological pathways. However, to search for rare variants in a systematic way, scientists need whole genome sequence data from thousands of people. The cost of obtaining these data was prohibitively expensive before the advent of ultra-high-throughput DNA sequencing 10 years ago.
No company was better positioned to leverage this technological breakthrough than deCODE. The support of the Icelandic nation and deCODE’s own unique capabilities have enabled it to assemble two massive data sets. One set contains genetic data provided by more than half the adults in Iceland, augmented by deCODE’s prowess in using genealogical records to predict genotypes. The other set, derived from Iceland’s universal health care system, contains phenotypic data on disease and other physical traits. By searching for non-random associations in these huge data sets, deCODE has found a wealth of rare and common genetic variants linked to disease risk.
deCODE’s mission of fundamental genetic research didn’t change with the acquisition by Amgen, and that focus has been enhanced by investments in next-generation sequencing technology. deCODE continues to publish its findings in top-tier science journals, including 28 papers in 2017.
Galaxies of Data
“Biology is choosing the targets for you”
Genetics and molecular genetics are not just a recent focus for Amgen. Seven of its medicines can be traced back to genes first cloned in the company’s labs. What’s different today is that Amgen is trying to go from one-off discoveries to an industrial-scale approach to gene-based research.
This more ambitious approach has its own distinct challenges. The targets that emerge may impact largely unexplored regions of biology, so teams pursuing these targets can’t rely on already published basic research. The targets themselves can be hard to drug because of their location or complex biology.
“With genetics, you’re not really choosing the targets—biology is choosing the targets for you,” Harper observed. “If you want to work in this space, you need capabilities that aren’t found in many companies. They include the ability to elucidate very complex biology, both in-house and through external collaborations. To interdict these targets you need a multimodality toolkit that gives you a lot of different therapeutic options. Amgen has always been strong in both of these areas, and we’re reinforcing these strengths. We’re investing in new technologies and expanding our talent pool by recruiting top scientists who are intrigued by the chance to work on ground-breaking research.”
Genetics-Driven Research: Benefits and Challenges
- More relevant to human biology than animal models of disease.
- Increased potential for large effect sizes.
- Higher clinical success rates.
- Increased potential for first-in-class therapies.
- Enhanced speed in some cases.
- Less need to validate target through assays.
- Confidence in target enables accelerated investment.
- Insights into safety and potential side effects.
- Targets may involve unexplored biology.
- Targets may be difficult to drug.
- Challenges with biology and druggability take time to solve.
- May be less informative for some therapeutic areas.
- Therapeutic hypotheses still need to be proven.
- Can pharmacology mimic gene variant present from birth?
- Not all variants provide both efficacy and safety insights.
One of those scientists is Wenjun Ouyang. He came to Amgen from Genentech two years ago and now serves as an executive director in the Oncology and Inflammation group. “One of the major reasons I joined Amgen was the genetics work with deCODE,” Ouyang said. “The pathways you identify from human genetics are already proven in humans, and the targets we find are worth the extra effort because the biology is so strong.”
In cancer research, the tools of human genetics have long been employed to identify tumor-suppressor genes and other targets useful in pinpointing tumor cells. Ouyang’s group is pioneering new ways to leverage sequence data through an approach called hypothesis-driven genetic research. Drug discovery scientists develop ideas for genes that might provide attractive targets, and data from deCODE is used to assess these hypotheses. Examples include an effort to find new cancer immunotherapies by identifying genes involved in autoimmune disorders.
“In these diseases, immune cells get hyper-reactive and start attacking organs and tissues,” said Ouyang. “Genes linked to these disorders may point to new pathways that regulate the immune response. We may be able to use these same pathways to activate immune cells to attack tumors.”
The Incredible Progress in DNA Sequencing
While numerous GWAS studies hint at where these new immune pathways might be found, GWAS data tend to be less precise and actionable, Ouyang said. “deCODE has helped us to validate GWAS data, pinpoint the specific genes involved, and identify new mutations in their own database. That enabled us to identify the best targets for new drug discovery programs.” This hypothesis-driven approach has already yielded multiple programs to date in both immuno-oncology and autoimmune disease. Amgen scientists working on other types of therapies are exploring the same approach.
CRISPR, IPSCs, and miniature organs
As Harper has noted, the pace of change in the biological sciences is accelerating. “The things scientists can do today were unimaginable just five or 10 years ago.” Human genetics exemplifies this rapid progress, and so do the tools now available to translate genetic insights into therapies.
Advances like CRISPR have made gene editing faster and easier, accelerating research needed to probe the function of newly discovered genes. Induced pluripotent stem cells (iPSCs) can theoretically be transformed into any cell type, including the cells that are most directly affected by key mutations. Major strides have also been made in constructing organoids, or miniature organs, which are useful in elucidating gene function at the level of human tissue. These and other emerging tools are helping Amgen build a more systematic approach to deciphering the biology resulting from gene variants.
Transforming Genes into Medicines
A similar revolution is underway in the realm of drug design, where Amgen’s investments have already given the company 13 modalities to choose from. “We’re not going to stand still in areas where the science is moving forward,” said Les Miranda, executive director for Hybrid Modality Engineering. “We’re prepared to explore new modalities to take on the targets we believe in.” For example, newer technologies like RNA interference can address some hard-to-hit proteins by intercepting the RNA used to make the proteins.
It takes patience to let the tools of drug discovery catch up with the biology revealed by human genetics. The reward is first-in-class molecules, higher odds of success in the clinic, and new therapies more likely to have the large benefits patients with serious illness are waiting for.
“Some companies in the industry have decided that real innovation is too high-risk,” said Harper. “We think it’s more risky to think you can simply buy all your innovation because too many companies are trying to do that. Our society will only pay for new treatments that make a major difference for patients, and to find these breakthroughs you need to be on the cutting edge of science. That’s where Amgen is now and where we intend to stay.”