A blog reader (thanks!) told me about a study published earlier this month in “Nature.” I’ll start with a quote from the Eureka Alert I read yesterday (http://tinyurl.com/5nesch): For the first time, scientists have decoded the complete DNA of a cancer patient and traced her disease – acute myelogenous leukemia – to its genetic roots. A large research team at the Genome Sequencing Center and the Siteman Cancer Center at Washington University School of Medicine in St. Louis sequenced the genome of the patient – a woman in her 50s who ultimately died of her disease – and the genome of her leukemia cells, to identify genetic changes unique to her cancer.
The scientists believe that these genetic changes originated from a single clone (!) and occurred one after the other—a sort of domino effect. They discovered three types of mutated genes: genes that are tumor-suppressors under normal circumstances, genes that promote cancer growth, and genes that may unlock the mystery of chemoresistance. Ahhh, chemoresistance…this phenomenon is unfortunately common to many cancers, including myeloma (incidentally, while doing research for this post I made an unexpected discovery…must do some major research this weekend…hope it pans out!).
The Eureka Alert provides a good, easy-to-understand summary. But I like to check my sources whenever possible, so I looked up the full study: http://tinyurl.com/57yd6j. Using a simple skin sample taken from the AML patient, these Washington University scientists found ten genes with acquired mutations. Two had already been studied, as they possess typical AML mutations and are involved in disease progression, but eight presented “new” mutations. The researchers determined that all of these mutations […] were present in nearly all tumour cells at presentation and again at relapse 11 months later, suggesting that the patient had a single dominant clone containing all of the mutations. No kidding: a single clone!!! Do I smell an AML stem cell, here?
The study is very detailed, very technical. At one point I found myself wading through a sea of alleles and genome gapped alignments and putative small indels and coding exons and split reads algorithms. Mamma mia. Eh, sometimes it is best to give up and glide gracefully over to a study’s Discussion part, which is usually more intelligible. So, let’s have a look at the Discussion. The fact that most of the genes discovered by this team have not been targeted in the treatment of AML justifies the use of next-generation whole-genome sequencing approaches to reveal somatic mutations in cancer genomes. I couldn’t agree more.
Of the unidentified (above-mentioned) eight mutations, four had not been previously implicated in cancer pathogenesis, which I found verrrry interesting. Further on we read that The importance of the eight newly defined somatic mutations for AML pathogenesis is not yet known, and will require functional validation studies in tissue culture cells and mouse models to assess their relevance. So mutations do exist, but they may be harmless, in other words. Probably not, though.
And also: the same mutations were detected in tumour cells in the relapse sample at approximately the same frequencies as in the primary sample. All of these mutations were therefore present in the resistant tumour cells that contributed to the patient’s relapse, further suggesting that a single clone contains all ten mutations. Ah, the single clone theory again.
The study ends on a note of prudence: For AML and other types of cancer, whole-genome sequencing may therefore be the only effective means for discovering all of the mutations that are relevant for pathogenesis. Okay, that, to me at least!, means that a cure for all of these presently incurable cancers, including my own, is not in the near future. This is not easy for me to say, but now I know why my skin crawls whenever I hear myeloma specialists declaring that a “cure” is around the corner or visible on the horizon (it just so happens that I recently heard one say words to that effect)…oh, I wish it were true…
But I don’t want to end on a negative note. I really do hope that genome sequencing for all cancer patients will shift quickly into high gear. And there are reasons to be optimistic. The sequencing technology has improved a great deal. It used to be very expensive and complicated to perform these genetic tests, plus the necessary genomic DNA samples had to be very large, but that is no longer the case. Reduced costs and smaller samples should make this new-generation technology more widely available to us.
It’s a first step.
It sounds like the Multiple Myeloma Research Consortium is
exploring the genetics of the disease….
Revolutionaries
Merck’s Free Radical
Matthew Herper 10.22.08, 6:00 PM ET
Forbes Magazine dated November 10, 2008
Stephen Friend
Cancer drugs don’t help 75% of the people who take them. Stephen Friend says he can use science to end the crapshoot
In the downtrodden drug industry, Merck cancer guru Stephen Friend may be one of the last great dreamers. His latest idea is one that would completely change the secretive and siloed way the pharmaceutical business fights cancer: create a giant, open-to-the-public database that will include every cancer drug and every patient and how that patient is doing. Track everything and over time we might be able to raise the abysmal success rate of treatment.
Friend, 54, has been a doctor who treated kids with cancer, an academic, an entrepreneur and a biotech chief executive. He helped develop a diagnostic test that predicts whether breast cancer will return after surgery. For five years he has been in charge of getting cancer drugs invented at Merck. Now 8 are in clinical trials, up from one, with 15 more preparing to enter trials. Friend is still unsatisfied. Why is it that, on average, three out of every four people who take a cancer medicine get lots of side effects but no benefit?
Researchers have been too willing to bet on hunches, he says, yet the technology to understand the complex biology of cancer is at hand. Spurred by Friend, Merck has spent billions on an arsenal of technologies for understanding how genes work. The resulting data stream is sent through the fastest supercomputer in the drug industry, a beast that consumes 64 kilowatts of power and is capable of 16 trillion calculations a second. Friend thinks he can accurately predict how groups of proteins in tumors work together and use that information to kill the cancer. He’s trying to drag the secretive world of drug-discovery chemistry into the computer age.
The Friend way would take all the data collected each year from the thousands of cancer patients entered in trials, make it anonymous and put it into one database, preferably held by the government but definitely accessible to any physician or scientist. Right now those data are lost to the wind once the trial is over. But by keeping track of patients’ genes, the genes in their tumors and what drugs they take, scientists will be able to discern patterns. Instead of trying drugs in order, from the ones that work most often to those that work least often, doctors will be able to pick the medicine that is most likely to help a particular patient. New medicines will get to market faster, along with diagnostic tests that will predict what will work. Friend predicts, somewhat optimistically, that prescribing decisions won’t be based on “a promotional campaign.” The database will decide.
“That future world is coming,” says Friend. “And pharmaceutical companies can live in that world. If you develop the best drug and develop it for the right patient, all this does is get it to that right patient.”
Merck has not done much so far to open its trial data to the world, nor have its rivals, but Merck has less to lose here and more to gain. It has fewer cancer drugs in human tests than Pfizer (nyse: PFE – news – people ) or AstraZeneca (nyse: AZN – news – people ), and its shares have dropped by half this year. Friend is powering ahead, building a first stab at the big database with the H. Lee Moffitt Cancer Center in Tampa, Fla. Over the next five years every patient who walks through Moffitt’s door will be asked to put genes and tumor samples in a database that will number 100,000 patients; 5,000 are already in. The database will provide information to the doctors doing research there and, eventually, to patients. If it turns out you have a gene that tells researchers what drug will work for you, Merck and Moffitt plan to let you know. Experiments that would have required weeks of thawing tumor samples now take a matter of hours.
“Right now most of medicine is based on a bunch of gray-haired guys who say, ‘This is the way I do it and it seems to work,'” says Moffitt Director Bill S. Dalton. “We need to determine over time what is useful and what isn’t. The only way to do that is to study 100,000 patients.”
The database idea is taking root elsewhere. The U.S. government is funding a Cancer Genome Atlas, in order to figure out how a large database would work. The Multiple Myeloma Research Consortium has funded the collection of 1,900 patients’ bone marrow samples that are being studied by the mit-Harvard Broad Institute, a genetic research center. New data from that effort will be available within months.
A megadatabase “could save me months or years of trying to collect patient information,” says Oregon Health & Science University oncologist Brian Druker, who helped get Novartis (nyse: NVS – news – people )’ potent tumor-fighter Gleevec to the market. But he questions whether researchers understand cancer biology well enough for Friend’s highly computational approach to pay off in the short term. “Over the long term the Merck strategy will be the winning strategy,” says Druker. “But right now I don’t think we’re quite there.”
Merck has spent the past few years trying to dig out of one of the toughest periods of its 120-year history. In 2003 several experimental drugs for various diseases failed, all at once. In 2004 the blockbuster painkiller Vioxx was yanked because it caused heart problems. Merck settled its Vioxx liability claims last year for $5 billion.
The stock recovered as eight drugs were approved in two years, but the revival was short-lived. Sales of its Vytorin cholesterol pill, produced with Schering-Plough (nyse: SGP – news – people ), have crashed under doubts about its effectiveness at preventing heart attacks. Cervical cancer vaccine Gardasil has hit a growth wall, and the Food & Drug Administration rejected a promising cholesterol drug because Merck had not collected enough safety data.
Merck hopes fighting cancer is one way out of this funk. Friend was put in charge of Merck’s cancer research efforts in 2003, two years after Merck bought the company he was running, Rosetta Inpharmatics. Friend had cofounded Rosetta in 1996 with Leland Hartwell, now director of the Fred Hutchinson Cancer Research Center in Seattle, and Leroy Hood, now president of the nearby Institute for Systems Biology. Like rival Affymetrix (nasdaq: AFFX – news – people ), Rosetta began selling tiny DNA chips that could be used to figure out how often cells were accessing their genes.
Merck bought Rosetta in 2001 for $620 million. Hood and Hartwell gave their shares to their institutions. Hartwell won the Nobel Prize six months later for other work. Friend made $10 million on the sale and built himself a solar-powered, off-the-grid house on Stuart Island.
The first fruits of Rosetta’s technology began to emerge with a 2002 article in the New England Journal of Medicine. Dutch researchers using Rosetta’s software found a particular pattern of genetic signals within breast cancer tumors that could predict whether or not the cancer would return after surgery. The test is not a significant product for Merck but was approved by the FDA in 2007. It and a similar test made by a rival, Genomic Health of Redwood City, Calif., are widely used to guide post-op treatment strategy.
Merck has been making big acquisitions to augment Friend’s technology. In 2006 Merck spent $1.1 billion in cash to buy tiny Sirna Therapeutics, a leader in a field called RNA silencing, which uses small molecules to shut off genes. These molecules can’t be used as drugs because the body destroys them. But they can be used in petri dishes to turn genes on and off to find out which are important.
This technology identified a gene last year called KRAS that predicts whether targeted cancer drugs like ImClone Systems (nasdaq: IMCL – news – people )’ Erbitux will work in a given cancer patient. Clinical trials confirmed this finding this year, and it turned out that 40% of the patients who were receiving Erbitux were getting no benefit. In the past this would have hurt the chances for a drug like Erbitux, but the new test makes doctors more eager to use the drug when it makes sense. Eli Lilly (nyse: LLY – news – people ) is now buying ImClone for $6.5 billion.
Friend has identified three families of cancer drugs that he thinks his technology can accurately understand: drugs that destroy DNA; those that mess up cell division; and drugs that block some of the most important signals in cancer cells. Noticeably absent are drugs such as Genentech (nyse: DNA – news – people )’s $2 billion (annual sales) Avastin, which stanches tumor blood supply. These are too complicated to understand.
He’s been buying the rights to drugs that fit his interests. In 2004 Merck bought Aton Pharmaceuticals for its drug Zolinza, used to treat cutaneous T cell lymphoma. In 2007 it pledged up to $1 billion for a cancer pill from Ariad Pharmaceuticals.
All of these bets are based on what Friend’s giant computer tells him. “This is going to have to be the path taken by pharma in the future,” says Hood of Friend’s current work. “It’s a gamble, but I think it’s one that if Merck sticks with it, they’ll win big.”
Recently Friend took a detour on his way to a research conference in Chicago. He flew to Florida, rented a 1972 Chevy Chevelle and drove to Cape Canaveral to watch the space shuttle launch. He says it wasn’t just that he wanted to recapture the feeling of the space race, when scientists were treated like heroes, but that he wanted to get a sense of a project that massive and complex. Creating a cancer drug is not that different.
“The puzzle’s gotten big,” he says of the cancer drug hunt. “But I think there is only one way to solve it.”
Sidebar:
Doctoring With Data
Subscribe to Forbes and Save. Click Here.
Your Rating Overall Rating
This is fascinating, Carla, many thanks for posting it.
Grazie!
Margaret