When Sean Harper took over as Amgen’s head of research in 2012, one of his first moves was to get the biotech giant to pay $415 million to buy DeCode Genetics, a struggling company in Iceland well-known for the huge DNA database it had built.
Today that bet paid off as scientists at DeCode described how they found a specific DNA defect that lowers the chance of having a heart attack by 35 percent, which they call the largest such effect ever found.
DeCode’s DNA detective work is “outstanding,” says Sekar Kathiresan, a scientist at the MIT/Harvard Broad Institute in Cambridge, Massachusetts. “This work immediately suggests that a medicine which mimics the protective mutation should reduce risk for heart attack.”
Amgen says it is already working on drugs that can copy the effect.
The discovery shows how some drug companies are exploiting gene databases to speed the search for drugs. Amgen’s Harper has said he believes the tactic could cut 18 months off the typical 14-year process of bringing a drug from the lab to market.
In their effort, DeCode searched the DNA of more than 300,000 Icelanders for unusual versions of genes previously linked to cholesterol levels. The gene variant they found occurs when one or both copies of a gene named ASGR1 are broken and don’t correctly make a protein. DeCode says about one in 120 Icelanders possess the gene variant and that, based on studies of their medical records, they have lower levels of bad cholesterol and live on average 1.5 years longer.
Kari Stefansson, the CEO and founder of DeCode, calls scouring databases in this fashion for protective genes a new way of creating drugs that isn’t based “on a beautiful hypothesis or lucky chemistry” but the facts of human biology as experienced by thousands of people.
The findings, described publicly today in the New England Journal of Medicine, were first made in July 2014 and since then Amgen has worked to further understand the biology of the gene and started developing candidate drugs. At a meeting with investors last year in which he touted the importance of DeCode’s work, Harper referred only to the discovery of a “Gene X.”
Amgen is betting that using DeCode’s database will slash the time it takes to develop drugs. That’s because the Icelandic scientists don’t only discover what biological process to tinker with, they can also help predict what the results of doing so will be long before any volunteer takes a drug.
What Amgen is searching for are versions of genes that protect against disease. In many cases, it happens that the DNA variations with the largest, most helpful effects actually destroy a gene’s function and stop it from making a protein (each of us is missing working copies of at least several of the total 20,000-gene complement that humans have). Most drugs also block proteins. What that means is that DeCode, which also has access to health records, can study walking, breathing examples of what someone would be like if they had taken a given protein-blocking drug their whole life.
That is hugely advantageous when compared to the usual way that companies test drugs—in animal models of human disease. But lab mice rarely predict how humans will respond. “The biggest problem in this industry is failure,” said Harper in an interview last year about DeCode. “If you can reduce that even incrementally, you can have a big impact.”
The Iceland database is among the most valuable because most of its citizens descend from a small number of original settlers who arrived there in the Dark Ages. That creates a “founder effect” in which some otherwise unusual genetic variants are much more common, allowing for analysis of their effects.
Amgen isn’t the only biotechnology company searching for genetic variables that may lead to good health. It directly competes with another biotech, Regeneron, which has its own mega-project to sequence the genomes of 100,000 people.
Harper, a doctor, joined Amgen in 2002. He rose to become R&D chief and took over a $4-billion-a-year R&D drug pipeline filled with risky bets and under criticism from Wall Street. Some analysts felt Amgen's research hadn't paid off and even asked if the company should be broken up or stop doing R&D.
Buying DeCode was part of a wider bet by Amgen that it can deliver the kind of breakthroughs the public always expected as the result of the human genome project. “What we are really interested in is very compelling evidence between human [DNA] variation and big, big risk of disease,” Harper says. He says DeCode has returned results “that you can really take to the bank.”
Gene breakthroughs are no guarantee of commercial success. A cholesterol-lowering drug that Amgen launched last year and which blocks a different gene, PCSK9, identified through older genetic research, registered sales of only $16 million in the first three months of this year.
Insurers have resisted paying for the drug, called Repatha, because it costs around $14,000 a year, and most patients can lower their cholesterol with $1-a-day drugs called statins.
That means that any drug to block ASGR1 will need to do more than simply lower cholesterol. Stefansson suggested such a drug could find use in families with a history of sudden death from heart attacks, or in others at elevated risk.
The database in Iceland isn’t useful only for planning the next blockbuster, however. When Amgen wondered if Repatha might cause cognitive problems, or serious side-effects like dementia, Stefansson was able to check the medical records of people lacking that gene, and report back no evidence of such a problem.
It can also help decide what drugs not to pursue. Harper says Amgen has already dropped plans for what could have been a $1 billion, 10-year study of another heart drug after DeCode’s data showed that people who lacked the targeted gene had lower cholesterol, but didn’t live any longer. That decision alone, says Harper, likely saved Amgen as much money as it spent acquiring DeCode.
How AI is reinventing what computers are
Three key ways artificial intelligence is changing what it means to compute.
These weird virtual creatures evolve their bodies to solve problems
They show how intelligence and body plans are closely linked—and could unlock AI for robots.
Our brains exist in a state of “controlled hallucination”
Three new books lay bare the weirdness of how our brains process the world around us.
We reviewed three at-home covid tests. The results were mixed.
Over-the-counter coronavirus tests are finally available in the US. Some are more accurate and easier to use than others.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.