The Birth of a Cancer Drug
Dr. John Mendelsohn is a very down-to-earth guy, more interested in scientific research than in getting his name in the society pages. But on May 22, the 64-year-old president of M.D. Anderson Cancer Center in Houston, one of the nation's top cancer hospitals, found himself hobnobbing with New York's social register at a lunch in his honor. About 100 of the city's upper echelon dined at the private home of Wall Street financier Martin E. Zweig, which just happens to take up the entire top floor of Manhattan's luxurious Pierre Hotel. Former President George Bush and his wife Barbara hosted the lunch, and before taking their seats the guests could check out Zweig's numerous Renoirs, his roomful of Beatles' memorabilia, and the famous crystal-studded dress that Marilyn Monroe wore when she sang "Happy Birthday" to President Kennedy in 1962--a dress that was auctioned off for $1.26 million two years ago.
Mendelsohn still managed to grab everyone's attention, however, with the description of a revolutionary cancer drug he created. The drug, dubbed IMC-C225, is the furthest along of a handful of new cancer treatments that precisely home in on a growth signal found in up to 50% of all cancer types. The week before the Pierre lunch, investigators reported on clinical trials of IMC-C225, combined with chemotherapy, at a major cancer research meeting. In those trials, the drug demonstrated remarkable success in causing colon cancer to regress in patients who had failed to respond to all other treatments. The drug also arrested growth in pancreatic, head and neck, and lung cancers, with the only side effect being an acne-like rash on the face. The lunch crowd was impressed enough by Mendelsohn's work to pony up more than $475,000 in donations to M.D. Anderson on the spot.
It was a generous pat on the back for the doctor, if a little late in the game. The 20 long years it has taken him to bring the drug from lab to market have been marked by a constant quest for funding. In fact, it is somewhat surprising that C225 still exists, given all the hurdles it had to overcome. From Mendelsohn's first efforts to convince someone, anyone, to finance his initial research at the University of California at San Diego in 1980, he struggled to persuade the cancer establishment that his targeted therapy would work. Twelve years later, when he finally found a company that believed in the drug--biotech startup ImClone Systems Inc. (IMCL ) of New York--he had to watch as the company and its two founders, scientist brothers with no business experience, struggled to raise development money.
By the time the Food & Drug Administration granted C225 fast-track approval status this past February, ImClone had flirted with bankruptcy, taken risks few companies could stomach, and poured $100 million into the drug's development. But unlike thousands of other drug candidates studied each year, C225 survived the process. ImClone's stock is now considered a biotech highflier and the company has a $3 billion market capitalization, even though it has never turned a profit in its 17-year history. In late June, ImClone started filing its request for FDA approval of C225 for colon cancer patients who have failed other treatments, with the official nod widely expected next February. Merrill Lynch & Co. analyst Eric M. Hecht predicts the drug, likely to cost $10,000 to $15,000 per patient, "will be a $500 million product in colon cancer alone," and could reach $1 billion in annual sales if used for a variety of other cancers.
What it won't be, despite all the hoopla, is a cure. Even in ImClone's pivotal colon cancer trial, only 22.5% of patients responded to the drug. But that is still the best response rate ever achieved with a drug in such a sick patient group. And C225 is just one of a slew of highly targeted cancer agents in the pipeline that could change the treatment model for this dreaded disease (table).
The watchword now in cancer treatment is not cure but control. C225 does not kill tumors--it stops them from growing. That's important, because a single tumor is not usually deadly. It is its ability to produce more tumors throughout the body that kills. Oncologists are hopeful that, in an ideal future, they will analyze a patient's tumor to determine what cellular mechanisms are driving the cancer's growth. Then, they can choose a drug designed to shut it down. The disease would become a chronic and treatable condition, much like diabetes. Patients could go on maintenance, receiving an intravenous dose of C225 or some other drug once every four or five weeks, possibly for years. "The idea of a cure may be overrated," says Dr. Harold Varmus, president of Memorial Sloan-Kettering Cancer Center in New York. "I would be quite content with just keeping the cancer under control. Patients can live long and healthy lives with a stable disease."
But if C225 is a harbinger of the future of cancer treatment, it is also a case study of how very hard it is to reach that future. There are currently 402 cancer drugs in development, compared with 126 a decade ago; how many will reach the market is anyone's guess. The rule of thumb in the pharmaceutical industry is that only one out of 5,000 drug candidates discovered in labs is commercialized. It seems human diseases have proven far more creative in evading drugs than drug companies have been in making them.
That's particularly true of cancer, which is really some 400 separate diseases. Each is driven by a different combination of cancer-causing genes and cellular growth signals. Even when a drug like C225 succeeds in breaking the code, its saga demonstrates that far more than good science is needed to guarantee success. It is often luck, timing, or a persistent scientist-champion that ultimately determines whether a drug survives the decade-long process of lab tests and clinical trials.
Mendelsohn is well suited to the role of crusader, having devoted his life to the treatment of cancer. As an undergraduate at Harvard University, he was the first student of Nobel laureate Dr. James Watson, a co-discoverer of the structure of DNA. From Watson he learned a love of molecular biology and the desire, as he puts it, "to use science to improve life." He went on to Harvard Medical School and in 1970 started conducting research into cancer at UC-San Diego.
"WAR ON CANCER." It was promising times for cancer research. In 1971, President Nixon declared a national "war on cancer" that led to the establishment of the National Cancer Institute to coordinate and encourage research. Millions of federal dollars were appropriated for the cause. Although Nixon's stated goal of wiping out cancer in just five years was wildly unrealistic, his funding push did engender tremendous progress in understanding the ways in which tumors develop and spread. A key breakthrough came in 1977, when Drs. Harold Varmus and J. Michael Bishop, then at the University of California at San Francisco, discovered that tumors are triggered by mutations in a small set of genes, called oncogenes, that control cell growth. Normal cells divide, replicate themselves, and die off millions of times over the course of a life. During the process, tiny mistakes are sometimes made and built into the oncogenes. These mistakes accumulate over decades until one day, the gene goes haywire. It releases a rush of growth signals that order the cell to start dividing and spreading very rapidly. Another group of genes, called tumor suppressors, is supposed to put the brakes on such wild cell growth, but they either fail or are overwhelmed.
In 1980, Mendelsohn surmised that a particular growth signal, known as the epidermal growth factor (EGF), would make a good drug target. EGF exists in small amounts in the epithelial cells that line our organs and make up skin. A rogue oncogene can trigger the production of large amounts of EGF, while at the same time causing the mutating cell's surface to sprout a million or more receptors designed to attract the growth factor. Once EGF binds to its receptors, it sets off a chain reaction of enzymes inside the newly cancerous cell that work for more, faster divisions. Thus, the cell stimulates its own rapid growth.
Mendelsohn knew that EGF receptors were found in large amounts in a number of cancers, including tumors of the head and neck, gastrointestinal tract, lung, kidney, breast, and prostate. He also knew that a cancer patient with a high degree of EGF generally had a very poor prognosis. Mendelsohn figured that blocking the receptor would stop the tumor from spreading.
The doctor asked the NCI to fund his research but was turned down, even though he had established the NCI center at UC-San Diego. The institute thought his research was too far out to fund without any supporting data. At the time, many scientists thought EGF would be a poor target because it is found in healthy as well as cancerous cells--raising the specter of significant side effects if the growth factor were blocked. But Mendelsohn believed healthy cells had other growth mechanisms that could compensate for the loss of EGF.
He finally managed to dredge up funding from private philanthropies. Mendelsohn and a colleague, Dr. Gordon Sato, along with a team of lab assistants, then spent two intense years developing a monoclonal antibody--an immune system protein created in the lab rather than in the body--that would attach to the EGF receptors before the growth factor could. "It's the same idea as sticking gum in a lock," he says. He reported the discovery of the antibody in 1983 and the following year announced that it stopped the growth of human tumors transplanted into mice.
SKEPTICISM. It was exciting work and much discussed among cancer researchers. But there was considerable skepticism about Mendelsohn's decision to use an antibody, which must be administered intravenously because it is too large a molecule to be absorbed as a pill. Even more worrisome, monoclonal antibodies at that time were made from proteins isolated from mice and so could not be tolerated for long by people.
Mendelsohn had won the NCI'S support, however. Once his research was published, the institute offered to finance the process of turning the mouse antibody into one with human protein so it could be more easily tolerated. By then it was the mid-1980s, and Mendelsohn had moved to New York to become chairman of the Medicine Dept. at Sloan-Kettering.
In 1988, Mendelsohn thought he had a promising cancer drug, but he needed someone to turn it into a commercial product. UC-San Diego owned the license to the original patents on C225 but was doing nothing to attract a licensee--and Mendelsohn had no interest in starting a business himself. "I wanted to do the research and let someone else raise the money," he says.
So the doctor started trolling the biotech byways himself, and eventually found Hybritech, San Diego's first biotech startup. Hybritech licensed the drug in 1990 and quickly found that C225 had a remarkable ability to eliminate tumors altogether in animals, especially when combined with chemotherapy and radiation. "You don't see that a lot in animals," says Dr. Roger B. Cohn, a cancer researcher at the University of Virginia Health Sciences Center. "Usually you just see tumors shrink. This was a particularly promising preclinical drug."
It seemed as though Mendelsohn could step back at this point and let Hybritech run with the ball--except that Hybritech sold the ball. In 1992, Eli Lilly & Co. (LLY ) acquired the startup and decided not to pursue C225. "It didn't fit in with their priorities at the time," says Mendelsohn. The license went back to UC-San Diego, and Mendelsohn once again faced the prospect that his discovery might never be used to benefit patients.
The fates intervened. One of Mendelsohn's good friends at Sloan-Kettering, Dr. Zvi Fuks, was on the scientific advisory board of a struggling biotech startup in lower Manhattan called ImClone Systems Inc. Fuks arranged a meeting in 1992 between Mendelsohn and ImClone's founders, brothers Samuel D. and Dr. Harlan W. Waksal, and the match clicked. "We decided to go full speed ahead and develop C225," says Sam, ImClone's chief executive. "Which wasn't easy because we didn't have a factory, we didn't have any clinical experience, we weren't prepared at all." They were, he now admits, "very naive."
Actually, everything about ImClone's beginnings had a naive, "gee, let's start a company" air about it. Sam, 53, and Harlan, 48, had no business experience when they founded the company in late 1984. Sam is a divorced man-about-town whose name shows up in the gossip pages of the New York tabloids. He's partial to well-tailored suits, lives in a Soho loft, and is good friends with Martha Stewart. Harlan, the chief operating officer, is more traditional, wearing shapeless sports coats and going home to his family in a leafy New Jersey suburb.
The brothers are very close, however. Their parents are Holocaust survivors--their mother was at Auschwitz and their father fought in the Polish underground. After World War II, the Waksals emigrated to Dayton, Ohio, where the father built a scrap-metal business.
At an early age, each boy turned to science. Samuel eventually got a PhD in immunology and worked for a while at the NCI, and Harlan became a pathologist. By the mid-1980s, both brothers were doing research at New York hospitals and living in the same building. "We started kicking around some fun ideas about how to start a company," says Harlan. "We thought we'd focus on infectious diseases, cancer, and diagnostics, make some products, get rich, and retire early."
It wasn't quite that easy. The Waksals raised an initial $4 million in venture capital and set up shop in a former shoe factory in a gritty part of lower Manhattan. They came up with the company name on a flight to Denver, where they were meeting with venture capitalists, by combining the names of the three businesses they thought they would be in: immunology, DNA cloning, and medical information systems. Their first projects were vaccines for sexually transmitted diseases, and they licensed some diagnostic technology from a Japanese company. In October, 1987, ImClone filed the papers for its initial public offering. A week later, on Oct. 19, the Dow Jones industrial average plummeted by 22% in a single day and the IPO was canceled.
Over the next few years, Sam engineered several funding deals to keep ImClone alive until the company went through with its IPO in 1991. ImClone still had not decided on a lead product, however, and its research was unfocused. The meeting with Mendelsohn couldn't have been timed better. The Waksals were determined to move the drug into human trials as quickly as possible and bought a bankrupt computer-chip manufacturer in Somerville, N.J. The plant was retrofitted for drug production and ImClone researchers began injecting the first cancer patients with C225 in 1994.
That's when the money dried up. Clinical trials are slow and expensive, and ImClone was rapidly running out of cash. The brothers had bet the farm on a drug few people outside the company believed in, and biotech stocks had crashed again in 1994. ImClone was trading at half of its IPO price of 13. Over a third of the workforce was laid off and others left, until only a bare-bones staff of 50 remained to keep things going. "We had zero cash--zero," says Sam. To stay afloat, ImClone spun off much of its other drug research, bringing in a paltry $7 million. Bankruptcy loomed large. "I was sure that something of great value wasn't going to be recognized because we didn't have the money," says Harlan.
During those dark days the brothers made a maverick decision: They would go it alone. Most biotech startups license their drugs or partner with a larger company that can underwrite the substantial development costs. In return, the bigfoot gets marketing rights--and most of the profits. Such deals have helped many biotech concerns survive, but they also limit any payoff. The prime example is Amgen Inc.: In its early days it sold most of the rights to its EPO anemia drug to Johnson & Johnson (JNJ ). Last year, J&J's EPO sales totaled $2.7 billion.
ENRAGED. The Waksals would have considered licensing C225 at first, but the offers that came in for their unproven drug were so low they became enraged. "We decided we would never give away the U.S. rights, no matter what. We knew this drug would work," says Harlan. They figured that if they could just stay afloat until the drug showed some promise in its first 12-patient, Phase 1 clinical trial, they would be able to raise more money. To get a sense of how big a risk they were taking, consider that Phase 1 studies are designed only to test a drug's safety, not how effective it is. Any beneficial results that arise are gravy.
The bet paid off. In the spring of 1995, C225 investigators reported that two patients with head and neck cancer who had failed all other treatments showed some tumor shrinkage. That May, Mendelsohn presented the data at the all-important American Society of Clinical Oncology (ASCO) annual meeting, one of the world's largest gatherings of cancer researchers. The stock started climbing, and big drug companies started approaching the Waksals, offering not only licensing proposals but takeover deals. The only deal ImClone made, however, was to license the European rights to the German company Merck KGaA (MKGAF ). "By now we knew how important C225 was," says Sam. They weren't about to step aside.
Mendelsohn, though, had stepped to the sidelines. He refused to conduct any clinical trials because he owned shares in ImClone (less than 1%) and thought it would be unethical. Besides, in July, 1996, he had taken over as president of M.D. Anderson, a mammoth hospital, leaving him little time to devote to C225. ImClone forged ahead, operating on a lean budget while slowly expanding its head and neck trials. "We would have done far more patients far more quickly if we had more money," acknowledges Sam.
ImClone might also have chosen a disease more common than head and neck. Colon cancer would have been a good candidate: It is the nation's third-most-common malignancy and the fourth-biggest killer. But ImClone's investigators felt that head and neck offered a better chance of proving that the drug worked, since those tumors express high amounts of EGF. The company could ill afford to pursue a less-than-sure thing.
Enter one very sick woman in Florida to change everyone's minds. Dr. Mark Rubin, an oncologist in Bonito Springs, Fla., diagnosed colon cancer in 28-year-old Shannon Kellum in 1998. A year later, Kellum's chemotherapy had failed, the cancer had spread to her liver and abdomen, and she was facing almost certain death. As luck would have it, Rubin had worked with Mendelsohn at Sloan-Kettering in the early 1990s, studying C225's effectiveness against colon cancer. He thought the drug would be worth trying on Kellum, and ImClone agreed to provide it on a compassionate basis. Kellum received the drug along with another round of the chemotherapy that had earlier failed.
The results were astounding. Kellum's tumors shrank by 80%, and the remainder were small enough to remove surgically. She remains free of the disease today. "She was a very brave young woman, and it was a very, very fortunate event," says Harlan. "Based on its value to this one patient, we decided to take a chance and test C225 on colon cancer."
THE ONE-TWO PUNCH. In October, 1999, ImClone set up a 125-patient study of the drug on colon cancer. The company chose only patients who had failed irinotecan, the best available chemotherapy, and so had only months to live. The patients would receive irinotecan again, but this time with C225, on the theory that the antibody would hold the cancer cells in check while the chemo poisoned them--a one-two punch. The lead investigator on the study was Dr. Leonard Saltz of Sloan-Kettering, a renowned colon cancer expert. "It was certainly gutsy of [ImClone] to do this, because it was a put-up or shut-up study," says Saltz. "There is a very high risk of failure in patients with proven resistance. If it didn't work, it could torpedo the whole program."
It worked. Early results released in November, 2000, indicated that at least 17% of the patients had tumor shrinkage of more than 50%. It was a strong enough response rate to prompt the FDA to grant fast-track status to C225 in February as a treatment for advanced colon cancer, guaranteeing a review of the drug's approval application within six months.
The FDA's confidence was not misplaced. When the trial ended last December, the tumors in 27 patients had shrunk by more than 50%, a 22.5% response rate--unusually high for such sick patients. The disease also stopped spreading in another nine patients. "I was hoping to see some activity, but this is more than I might have expected," says Saltz.
The 2001 ASCO meeting in San Francisco was C225's coming-of-age party. The Waksals even threw a bash for attending researchers, featuring 1970s rock band the Doobie Brothers. ImClone presented data at the meeting not only from the colon cancer study, but pancreatic and head-and-neck cancer trials as well. The pancreatic cancer study got particularly close attention from oncologists because the disease is virtually untreatable and can kill within months. In a Phase 2 trial, 40 patients in advanced stages of the disease were given chemotherapy and C225. The tumors shrank by more than 50% in five patients, and another 16 either stabilized or had a partial response. Overall survival rate at one year was 32.5%, compared with 18% on chemotherapy alone.
Those numbers still aren't great. Sadly, it is the nature of cancer studies that a success can mean nothing more than adding a few months to a patient's life. For desperately ill patients seeking any reprieve they can get, this may not matter. Shirley S. Heligman, a 73-year-old Manhattan resident, was diagnosed with very advanced metastatic colon cancer in 1998 and had failed all other treatments when she enrolled in Saltz's C225 trial in January, 2000. "It was the best thing that happened to me since the diagnosis," she says. "It held the cancer at bay for 10 months, which in my stage of disease is fantastic." C225 did eventually stop working, and Heligman is back on traditional chemotherapy, but her determination is strengthened. "Hopefully, they will keep coming out with something new, and I will try them all."
Researchers believe that C225 and other targeted therapies could have a much stronger impact if given earlier in the course of the disease. ImClone is conducting clinical trials to test that theory for head and neck cancer and is planning similar trials for colon and pancreatic cancer. Mendelsohn hopes there will eventually be extended clinical trials to determine how long a patient can remain on the drug without the cancer flaring up again.
The Waksals, though, have a more immediate challenge. Their next trial by fire will be making and marketing the drug. Few small companies have the resources or the expertise to handle this stage alone. The manufacturing process, for example, must undergo tough FDA scrutiny that can often delay a drug. Just recently, the agency announced a delay in the approval process for an important new sepsis drug from Eli Lilly because of manufacturing issues.
NO GUARANTEES. ImClone is taking a green-fields approach, building a new, $50 million plant in New Jersey rather than retrofitting an old one, which analysts say should make the approval smoother. The plant is scheduled to start up in August, a rapid nine months after construction began. The company is planning to break ground in October for a second plant that is to have three times the capacity.
On the marketing side, the Waksals scored a hiring coup in 1998 when they lured 38-year-old biotech marketing whiz Ronald A. Martell away from Genentech Inc. Now ImClone's vice-president of marketing, he engineered the successful marketing campaign for Herceptin, Genentech's (DNA ) top-selling antibody for breast cancer. Waksal says the company will likely hire about 70 salespeople as the drug moves closer to FDA approval--but it may not need to. Big pharmaceutical companies often buy up biotech concerns once the research is done, allowing them to skip the risk and leverage their own considerable marketing organizations. Sam insists the company is not for sale, but takeover speculation is already swirling around the company. "This drug would make an excellent platform for a big pharma company that wants to establish a major presence in cancer," says Merrill Lynch's Hecht. "A year from today [ImClone] probably won't be around anymore."
Whoever owns C225 in a year will soon be facing competition. Several other drugs that target EGF are now in clinical trials, led by AstraZeneca's (AZN ) Iressa and OSI Pharmaceuticals' (OSIP ) OSI-774, both of which are in easier-to-take pill forms. However, these medicines, which enter the cell rather than bind to its surface, carry greater risk of side effects, including diarrhea. That could make them unsuitable for colon cancer. Many cancer experts also believe the pills and antibodies could complement each other, allowing oncologists to tailor treatment to the needs of the individual patient.
Much could still go wrong, of course. There is no guarantee that the FDA will approve the drug in February. The agency often asks for more data, adding many months to the process. As time goes by, C225 may not live up to its early promise. Some disappointment is already gathering around Gleevec, the leukemia drug from Novartis (NVS ) that was approved by the FDA in May in a record 2 1/2 months after filing. Although the drug halts the progress of this rare and deadly blood cancer in some 90% of patients, continuing trials found that the sickest patients eventually develop resistance.
Still, desperately ill cancer patients like Heligman are eager for anything that might win them more time. Targeted cancer drugs are the best option they have been offered, and cancer researchers are galvanized to find more and better ones. "For the first time, we may be able to say this is the beginning of the end of chemotherapy," says Dr. Larry Norton, a leading cancer researcher at Sloan-Kettering. If so, it will be truly a sweet reward for Mendelsohn and the many other scientists that struggled to make it so.
By Catherine Arnst